Sensory Features in Autism: Physiological and Behavioral Characterization

Background: Sensory hypo and hyper-responsivity, and unusual sensory interests are extremely prevalent (80-90%) in individuals with Autistic Spectrum Disorder (ASD) and present some of the most challenging obstacles by limiting adaptive behaviors and participation in life activities. They are now proposed to be included as a core feature for diagnosis of Autism Spectrum Disorders (ASD) under the Restricted and Stereotypic Behavior criteria in the DSM 5 (APA, 2012).  As such, objective markers of these features have new and important relevance for diagnosis and treatment of ASD.  

Objectives:  This paper reports on the results of an NIH-funded study that evaluated physiological and behavioral responses to sensation in 59 phenotypically characterized subjects with ASD, 6-9 years of age in comparison to 30 age and IQ matched controls.  The aims of this study were to 1) evaluate sympathetic and parasympathetic activity at baseline and during sensory challenges in comparison to typically developing controls to determine if physiological activity during sensory stimuli is a unique feature of ASD; and 2) evaluate whether physiological reactions to sensation are related to (or predict) behavioral responses to sensation and/or adaptive behavior.

Methods: Fifty-nine children diagnosed with ASD and confirmed with the ADI-R were tested during the Sensory Challenge Protocol, a unique laboratory procedure designed to assess autonomic nervous system activity in response to sensory challenges in the auditory, tactile, olfactory, visual and vestibular systems. Parasympathetic activity was measured via heart rate variability, and sympathetic activity was measured via pre-ejection period as described by Berntson, Cacioppo and colleagues (1995).  Behavioral responses to sensation were measured by the Short Sensory Profile and the Sensory Processing Measure; and The Vineland Adaptive Behavior Scales-II were used to assess adaptive behaviors. Mixed effects linear regression was used to jointly model RSA and PEP scores at each domain by group.  Within the mixed effects model, we performed several multivariate hypothesis tests.  First, we tested for any difference between groups at any domain with respect to mean RSA/PEP.  Second, we tested for any difference between groups with respect to change in RSA/PEP from the previous domain. If the multivariate tests were significant, we proceeded to perform group comparisons at each domain.  In addition, multiple linear regression analyses estimated the association between behavioral scores (the dependent variables) and physiological measures controlling for severity of ASD, gender, and mental age and IQ.

Results: Subjects with ASD show significantly less parasympathetic reactivity during the sensory challenges and a trend for elevated sympathetic activity.  Data analysis of physiological-behavioral relationships is currently being completed and will also be reported.  We expect that decreased parasympathetic activity and increased sympathetic activity will be related to greater sensory dysfunction and poorer adaptive behavior.   

Conclusions: Objective characterization of sensory features in ASD may provide important clues regarding the mechanisms of these sensory features and provide insight into intervention targets for these behaviors.