Note: Most Internet Explorer 8 users encounter issues playing the presentation videos. Please update your browser or use a different one if available.

The Physiological Correlates of Anxiety in Autism Spectrum Disorders

Saturday, 4 May 2013: 09:00-13:00
Banquet Hall (Kursaal Centre)
M. J. Hollocks1, L. E. Grayson1, A. Papadopoulos2, P. Howlin3 and E. Simonoff1, (1)Child & Adolescent Psychiatry, Institute of Psychiatry, King's College London, London, United Kingdom, (2)Psychological Medicine, King's College London, London, United Kingdom, (3)Department of Psychology, Institute of Psychiatry, King's College London, London, United Kingdom

People with ASD are known to display increased levels of anxiety symptoms and disorders. However the physiological correlates of this vulnerability to symptoms of anxiety are yet to be fully explored. 


To determine whether people with ASDs show a similar physiological response to psychosocial stress as measured by salivary cortisol and heart rate (HR) compared to controls, and whether physiological response relates to anxiety.  


This study includes 53 boys with ASDs and 23 typically developing controls all participants were male and their full-scale IQ was ≥70. We specifically recruited ASD participants with and without anxiety problems.  

Anxiety was assessed using the Spence Child Anxiety Scale (SCAS) – parent version. For the present analysis, the ASD group was collapsed and anxiety symptoms treated as an independent measure. Participants underwent a psychosocial stress test consisting of 40 minutes rest, 20 minutes stress and finally 40 minutes recovery. The stress test comprised a non-verbal drawing task and a public speaking task.  Six cortisol samples were taken, at baseline, and then at 20 minute intervals.  HR was measured continuously and divided into five 20 minute segments.


The groups differed in age (ASD 12.8 years vs. controls 14 years (p ≤ .05)) but not FSIQ (102 vs. 116), as expected by design, the ASD group had significantly higher anxiety ratings (mean SCAS-P; 33 vs. 9, p ≤ .01). 

Cortisol. A repeated measures ANOVA revealed significant main effects for group (p= .02), time (p ≤ .01) and a groupxtime interaction (p= .03).  The cortisol concentration was significantly lower in the ASD group at each post-stress time-point (all p ≤ .01).  

Mean Heart Rate. A repeated measures ANOVA also revealed significant main effects of group (p = .02) and time (p ≤ .01) and a significant groupxtime interaction (p = .011). The ASD group mean HR was significantly higher in the resting phase and in the first 20 minutes of the recovery phase (both p ≤ .01), but did not differ in the stress phase.

Stress responsiveness.  The ASD group showed reduced stress responsiveness compared to controls as demonstrated by reduced cortisol (.54 versus 2.33 (p=.011)) and a less marked increase in HR 5.4 beats-per-minute versus 10.4 (p ≤ .01

A regression analysis to examine the independent roles of anxiety and participant group on physiological responsiveness revealed that increasing SCAS score (p=.06) but not group predicts a decreased cortisol response to social stress.  

Analysis of HR responsiveness revealed a significant SCAS by group interaction (p=.012). In the control group, increasing SCAS scores were positively associated with increased stress reactivity while the opposite relationship was seen in the ASD group. 


Children with ASD displayed elevated resting HR and a lack of responsiveness in both HR and cortisol when entering a stressful situation.  Furthermore the relationship between stress responsiveness and baseline anxiety symptoms appears to differ between groups. In those with ASD reduced physiological response to stress is associated with greater levels of anxiety. Results will be discussed in the context of heart rate variability.

| More