Relationship Between Sensory Over-Responsivity and Anxiety in Young Children with Autism Spectrum Disorders:

Background: Anxiety disorders are extremely common in children with autism spectrum disorders (ASD), and can increase the functional impairment of these children (Lainhart, 1999), with rates two to three times as high as in the general population (Sukhodolsky et al., 2008).  Sensory over-responsivity (SOR) is another common, and impairing feature found in more than half of children with ASD, and SOR has been linked to anxiety in children with ASD (Ben-Sasson, Cermak, Orsmond, Tager-Flusberg, Kadlec, & Carter, 2008; Liss, Saulnier, Fein, & Kinsborne, 2006; Pfeiffer, Kinnealey, & Herzberg, 2005). While there is some evidence that anxiety and SOR are associated in children with ASD, it is unclear what reciprocal patterns of influence (if any) exist between the two.  A greater understanding of patterns of influence over time has direct implications for etiological study and intervention choices in this population.

Objectives: The purpose of this investigation was to gain a better understanding of associations between anxiety and SOR in children with ASD by examining the extent to which anxiety predicted changes in SOR and vice versa in young children with ASD. 

Methods: Participants were 149 toddlers with an ASD diagnosis who were between 18 and 33 months at initial assessment, and between 28 and 50 months at a follow-up assessment approximately 1 year later.  Anxiety and sensory sensitivity (SOR) were assessed by mother’s report on the Infant-Toddler Social and Emotional Assessment (ITSEA) at both time points.  Autism symptoms were assessed at time 1 using a combined social-communication score on the Autism Diagnostic Observation Schedule (ADOS).

Results: Anxiety and SOR were significantly correlated at time 1 (r=.52, p <.001) and time 2 (r-.60, p<.001).  Mean anxiety scores increased from time 1 (M=.31, SD=.24) to time 2 (M=.35, SD=.30), t(148) = -3.00, p=.003.  However, mean SOR scores were stable from time 1 (M=.53, SD=.43) to time 2 (M=.57, SD=.43), t(148) = -1.10, p=.275.  Cross-lagged panel analyses were conducted in Mplus to compare the two pathways (SOR to anxiety and vice versa) controlling for the autocorrelation within measure.  Time 1 child age and autism severity as measured by the ADOS social-communication score were entered as covariates.  Results indicated a significant cross-lagged effect between time 1 SOR and time 2 anxiety after controlling for time 1 anxiety, child age, and autism severity (B=0.13, SE = .05, p=.01).  The reciprocal cross-lagged effect from time 1 anxiety to time 2 SOR was not significant (B=.19, SE=.14, p=.18).

Conclusions: The presence of SOR appears to contribute to changes in anxiety, which overall increased in this group of children with ASD.  In contrast, the early presence of anxiety does not appear to contribute to changes in SOR, which was quite stable. These results support SOR as a possible contributor to anxiety in children with ASD.  Implications for further study of the etiology of affective disorders and intervention choices in this population will be discussed.