The literature on sensation in Autism Spectrum Disorders (ASD) is sparse and empirical data related to sensory processing impairments in ASD is lacking. The few studies examining sensation do not use objective measures of sensory reactivity. Therefore, this study was implemented to objectively evaluate the pervasiveness and variability of sensory dysfunction in children with ASD.
The four aims of this study were to evaluate: 1) the reliability of the physiologic laboratory paradigm the Sensory Challenge Protocol in school-aged children (ages 5-15) with ASD; 2) the variability of sensory reactivity among children with ASD; 3) the relations among sensory symptoms and core symptoms of autism (e.g., social behaviors, communication, repetitive behaviors); and 4) the convergent validity of the laboratory measure with parent report measures of sensory symptoms and adaptive behavior.
Thirty-eight children with ASD, ages 5-15, participated in this study. Children were enrolled in the study by the Autism and Developmental Disabilities Clinic at the University of Colorado Health Sciences Center. All met the criteria for Autistic Disorder or Asperger Syndrome based on the Autism Diagnostic Observation Schedule and the Autism Screening Questionnaire. Fourteen of the 38 subjects participated in the test-retest study conducted 4 to 6 weeks after the initial session.
The following behavioral measures were administered to each participant: the Child Behavior Checklist, the Carey Temperament Scale, the Leiter International Performance Scales-Revised IQ, Parent rating and Examiner rating, the Short Sensory Profile, and the Vineland Adaptive Behavior Scales. Physiologic data were also collected on each participant during the administration of our laboratory paradigm, The Sensory Challenge Protocol. This paradigm records resting physiologic arousal and physiologic reactivity after sensation using two objective measures, electrodermal activity (EDA) and vagal measures (VM). EDA and VM are recorded continuously during a 3 minute baseline period with no stimulation, followed by the presentation of eight trials in each sensory domain (tone, visual, auditory, olfactory, tactile, and vestibular). The experiment ends with a 3 minute recovery period during which no stimuli are delivered.
The results of this study confirm the reliability of the physiologic measures. Both tonic and phasic measures were useful in characterizing children with ASD. Baseline physiologic arousal clearly separated our sample in to two groups and the baseline tonic arousal predicted sensory reactivity, the phasic reactions to specific stimuli. Children in the high arousal group tended to have larger and faster reactions to the sensory stimuli and were slower to habituate during the repeated trials. Behavioral and physiological heterogeneity was present within both diagnostic groups. Within each diagnostic group there was a wide range of scores including children who were more reactive and others who were much less reactive. Behavioral heterogeneity of sensory responsivity was supported by the Short Sensory Profile factor analysis. Across diagnostic groups, sensory functioning clustered into three groupings, Sensory Over-Responsivity, Sensory Under-Responsivity and Sensory Seeking. This finding confirms our working conceptual model of Sensory Modulation Disorder which we hypothesize exists in many disability groups.
A consistent association was obtained between sensory seeking behaviors as measured by the Short Sensory Profile and one of the core symptoms of ASD, repetitive behaviors. This association was strongest in the Asperger group, the low arousal group and the group that habituated to tone. In addition, when children with Autism had more core symptoms, they also had higher physiologic reactivity; children with Asperger Syndrome tended to have lower physiologic reactivity if they had more core symptoms. These results are intriguing and should be further evaluated in future studies.
The concurrent validity of the physiologic measures is supported by the associations with the behavioral measures. Most noteworthy was the strong relation that was obtained between vagal tone and physiologic reactivity and between vagal tone and functional aspects of daily living as measured by the Vineland Adaptive Behavior Scale. This finding suggests two important implications: 1) children who have difficulty regulating their arousal have greater problems in adaptive behavior and 2) an optimal level of reactivity is necessary to maximize the performance of daily living skills.
Temperament was differentially associated with the sensory characteristics of the two diagnostic groups. For children with Autism the physiologic measures were more strongly associated with an inhibited, anxious behavioral style and for children with Asperger Syndrome the physiologic measures were more strongly associated with measures of a disinhibited, impulsive behavioral style. Stronger correlations were also obtained for the Asperger group between the physiology and problems in sensory processing. Most notable were problems in tactile sensitivity, taste/smell sensitivity m visual auditory sensitivity and low energy/weak.
This study has advanced the literature on children with ASD by describing the range and variability of sensory processing difficulties in this population and demonstrating the usefulness of an objective laboratory measure of sensory reactivity. Children with Autism and Asperger's Syndrome demonstrated significant symptoms of sensory processing disorder that were documented physiologically as well as behaviorally.
We propose new directions for future data analysis. Visual inspection of many scatter plots suggests that there may be three groupings of functioning of physiologic variables. The factor analysis of Short Sensory Profile items also suggests three behavioral categories. Additional empirical analyses of existing data using both the three physiologic groups and/or the three sensory groups should be illuminating. This will further clarify the role of sensory processing in Autism and Asperger Syndrome and could significantly impact the diagnosis and treatment of children with ASD. Replicability of these findings with a larger sample is essential to cross-validate our preliminary results.