How Can We Reliably Measure Sensory Reactivity?

Background: Sensory processing difficulties are common across neurodevelopmental disorders, and have gained increasing attention in autism since they were added to the diagnostic criteria in the DSM-V. However, the proliferation of sensory-related studies in autism has highlighted the need for reliable behavioral measures of sensory reactivity. This is especially true given that one of the most commonly used measures, the Short Sensory Profile, a parent report questionnaire, has recently been shown to have poor psychometric properties in autism (Williams et al., 2018). This has motivated the development of better clinical and research measures to characterize sensory processing. Specifically, previous research suggests that sensory over-responsivity (SOR), an adverse response to sensory stimuli, is one of the most common and impairing sensory processing difficulties in autism and occurs most frequently in the auditory and tactile domains (Tavassoli et al., 2016).

Objectives: Hence, this study seeks to advance the field of sensory assessments by comparing the auditory and tactile over-responsive items for children with neurodevelopmental disorders using parent report and direct assessment (SP-3D:A), and by providing cut-off scores.

Methods: This study included 304 children from a mixed neurodevelopmental disorders cohort including autism and sensory processing dysfunction (n= 176) as well as neurotypical children (n= 128). The widely used Short Sensory Profile caregiver report was used, in which parents rate how often their child shows a particular sensory behaviour. The SP-3D:A, a direct structured observational tool, which is measuring behavioral response to specific sensory stimuli, was also utilized. The SP-3D:A is ideally suited for this task, as it includes characterizations of SOR in both auditory and tactile domains. To achieve an SOR specific score for the auditory and tactile domains, we chose items on the SSP and SP-3D:A that are reflecting SOR behaviors by clinical consensus. For each measure, we then calculated the TDC mean +1 SD to create cut-scores.

Results: Using direct SP-3D:A assessment, 31% of the children with neurodevelopmental disorders (NDD) had AOR and 27% had TOR. Using the SSP parent report, 62% of the children with NDD were classified as having AOR, whereas 68% had TOR. The Inter-test-agreement between SSP and SP-3D:A for AOR was 65% and TOR was 50%. Moreover, as expected the NDD group was significantly more affected by SOR than the TDC group (χ2≥17.5, p≤.0001).

Conclusions: This study identified cut-off scores for AOR and TOR using the SSP parent report and SP-3D:A observation. We found that the parent report questionnaire and direct observation have only a moderate overlap. A combination of questionnaire and direct observation measures should be used in clinical and research settings; specifically, the use of sensory questionnaires as a screening tool, followed by standardized direct observations. In line with previous reports, we find that more children meet SOR criteria based on parent report than on direct assessment in both the auditory and tactile