Stability of Autonomic Arousal and the Relationship with Cognitive and Social Skills

Background: There is an increasing interest in physiological biomarkers (such as heart rate (HR) and high-frequency heart rate variability (HRV)) in Autism Spectrum Disorder (ASD). However, we have little information about the reliability of autonomic arousal in children with ASD. Moreover, studies have reported contrasting results when it comes to group differences, with support for lower parasympathetic activity in ASD or no group differences compared to typically developing (TD) children. Also, the relationship between autonomic arousal and developmental abilities is unclear. Within TD children, lower HRV and higher HR during periods of rest has been associated with poorer developmental outcome. In ASD similar results are found in middle-childhood, but not yet earlier in development.

Objectives: This study has three aims: 1. To compare toddlers with ASD and TD on HR and HRV; 2. To look at measurement stability using intra-individual variability and test-retest reliability, 3. To examine the relationship between autonomic arousal and cognitive and social abilities.

Methods: Heart rate was collected during two visits (M = 19.2 days apart) for 71 children with ASD and 67 children with TD (2 to 4-years-old). Social and cognitive abilities were measured using the Preschool Language Scale, the Communication and Socialization subscales of the Vineland Adaptive Behavior Scales, and the preschool Behaviour Rating Inventory of Executive Functioning. The children watched four 90-second “rest” videos (wildlife animals) as part of a screen-based battery. Data was analysed in 30-second epochs. First, average HR and HRV between groups was compared using t-tests. Second, stability of the physiological measures was evaluated (a) by calculating intra-individual variation and (b) with interclass correlations. Third, we ran separate regression analyses with HR and HRV as the independent variables and Language, Communication, Socialization and EF as the dependent variables, controlling for age and Visual Reception score of the Mullen Scales of Early learning.

Results: Groups did not differ in HR (t(136)=-1.33, p=.187, d=-.23) or HRV (t(136)=0.87, p=.386, d=.15). Intra-individual variability of HR was higher in the ASD group than the TD group. For HRV, this was only the case for the second visit. Test-retest reliability of HR and HRV was average to good in both groups (ICC range: .59-.73). There was an interaction effect of HR & group and HRV & group for Language, Communication and EF, but not Socialization. The effect on Communication remained after controlling for age and visual reception. Post-hoc analyses showed that only in the ASD group there was a positive correlation between HRV and Communication (partial r =0.30, p=.013).

Conclusions: First, in this large sample, measures of autonomic control do not discriminate between 2 to 4-year-old children with ASD and TD during rest. Secondly, test-retest reliability was comparable in both groups; however, the ASD group showed higher intra-individual variability within a session. HRV was positively correlated with communication abilities in the ASD group, but not the TD group. Overall, these findings suggest there is an association between autonomic arousal and communication skills in ASD; variability within a person is important to consider in future research.