Examining the Relationship between Early Sensory Reactivity and Later Autism Severity

Background: Infants at high-risk (HR) for developing Autism Spectrum Disorder (ASD) often display atypical patterns of sensory reactivity, including hypo-reactivity, hyper-reactivity, and sensory seeking behaviors. Sensory seeking behaviors often increase over the second year of life in children later diagnosed with ASD (Baranek et al., 2018) and hypo-reactivity to sensory stimuli at 24 months has also been linked to later ASD diagnoses (Germani et al., 2014). Extending this work to younger samples and evaluating how sensory reactivity changes over the first years of life has implications for early identification and treatment planning.

Objectives: To examine the relationship between sensory reactivity at 12 and 24 months and later ASD severity.

Methods: This study includes 87 HR children (60 boys) identified at 12 months based on community screenings with the First Year Inventory 2.0 (Baranek et al., 2003). Children completed the Sensory Processing Assessment (SPA; Baranek, 1999) at 14 months (+/- 0.77; Time 1) and 23 months (n=83; +/- 0.86; Time 2). 50% (n=45) of the children were randomly assigned to an intervention trial. A subsample (n=45) returned between ages 3 and 5 years (mean= 4.5 years +/- 0.92; Time 3) for follow-up diagnostic assessments and received the Autism Diagnostic Observation Schedule (ADOS-2; Lord, Luyster, Gotham, & Guthrie), yielding calibrated severity scores: overall score (CSS), CSS social affect (CSS SA), and CSS restricted, repetitive behaviors (CSS RRB). To account for missing data, multiple imputation was carried out using Multivariate Imputation by Chained Equations (see MICE package in R). Linear regression models were fit to the imputed datasets with each SPA score at Time 1 and the change in each SPA score from Time 1 to Time 2 (covariates: treatment group and age at Time 3). Effect sizes were calculated (the parameter estimate divided by the standard error).

Results: SPA hypo-reactivity at Time 1 was significantly related to CSS [95% CI (0.35, 3.43); effect size = 2.49] and CSS SA [95% CI (0.33, 3.32); effect size = 2.48]. Increases in SPA Hypo-reactivity from Time 1 to Time 2 was significantly related to CSS [95% CI (0.41, 2.78); effect size = 2.73], CSS SA [95% CI (0.30, 2.61); effect size = 2.57], and CSS RRB [95% CI (0.05, 3.03); effect size = 2.11]. Increases in SPA Hyper-reactivity from Time 1 to Time 2 was significantly related to CSS [95% CI (0.45, 4.79); effect size = 2.46], but not CSS SA or CSS RRB. Sensory Seeking behaviors at Time 1 and changes from Time 1 to T2 were not significantly related to later ASD severity.

Conclusions: Child hypo-reactivity at 12 months may be particularly helpful in predicting ASD severity during the preschool years. In addition, increases in hypo- and hyper-reactivity from 12 to 24 months may provide new information about the emergence of ASD symptoms prior to a diagnosis. These results indicate that early sensory reactivity may be a useful precursor to an ASD diagnosis and may provide a new target for early interventions.