Resting Autonomic Activity in Children with Autism Spectrum Disorders and Matched Controls with and without Comorbid Anxiety

Background: Anxiety disorders are common among children with autism spectrum disorder (ASD). Both anxiety and ASD are associated with abnormal physiological activity. To date, few studies have investigated the autonomic profile of participants with ASD at rest and none has systematically assessed how the co-occurrence of formal anxiety disorders and ASD modulates resting physiological activity.

Objectives: The aim of the present study was to evaluate sympathetic and parasympathetic activity at rest in children and adolescents with ASD with and without comorbid anxiety. We also aimed to compare their autonomic profile, as well as parental and subjective reports of anxiety symptomatology, with the profiles of typically developing controls (TDC), matched for gender, age and presence of clinically relevant anxiety.

Methods: ASD was determined based on results from the ADOS, developmental interview or ADI-R, and clinical judgment. Both ASD and TDC participants received the Anxiety Disorders Interview Schedule (ADIS). Based on these data  4 subgroups were identified: ASD with comorbid anxiety (ASDanx, N=22), ASD and no anxiety (ASDnon, N=15), TDC with comorbid anxiety (TDCanx, N=16) and TDC and no anxiety (TDCnon, N=22). Participant ages ranged from 8 to 17 years old and were comparable across groups. In only the ASDanx group, males significantly outnumbered females. Autonomic parameters were assessed during a 5-minute resting state collected while the child was watching neutral nature scenes. Skin conductance level (SCL), a stable proxy for sympathetic activity in children with ASD (Schoen et al. 2008), and  the proportion of high frequency component of heart rate variability (HRV), a proxy of parasympathetic activity, were calculated. Parents and children were asked to complete the Screen for Child Anxiety Related Emotional Disorders (SCARED); the overall score was used as a subjective evaluation of anxiety.

Results: Reduced sympathetic and parasympathetic activity at rest was found in ASD as compared to TDC. In particular, the analysis of sympathetic activity revealed a paradoxical hypo-arousal in ASDanx participants. Moreover, the analysis of the parasympathetic cardiac activity showed that in TDC (but not in ASD) anxiety promotes tachycardia. Overall, ASD groups showed lower parasympathetic activity, in line with the idea of inflexible physiological reactivity. Regression analyses suggest that the concurrent consideration of physiological responses as well as parent, and self-reports allows for an efficient categorization of the 4 subgroups. 

Conclusions: Anxiety in ASD was associated with abnormal sympathetic and parasympathetic activity which, in association with parent and self-reports, allowed for efficient diagnostic categorization. As expected, the TDCanx group showed elevated levels of sympathetic activity. On the contrary, children with ASD and comorbid anxiety presented an atypically blunted autonomic profile. This suggests that anxiety in ASD may be related to different mechanisms than in TDC. In addition, the combination of autonomic profile and anxiety report measures might constitute a marker of anxiety in ASD. These findings emphasize the need for further investigations of the physiological activity of ASD and anxiety, since identification may facilitate the selection of effective treatments.