Atypical Visual Attention and Autism Severity in Infants at High Risk for Autism Spectrum Disorders

Background: Early visual attention control is a foundational developmental skill important for social engagement in infancy across typical and clinical populations. Evidence suggests that individuals with Autism Spectrum Disorders (ASD) have a primary impairment in social functioning shown to manifest in early infancy with reduced eye contact, atypical gaze behaviors, and atypical attention disengagement (Elsabbagh, 2013). Furthermore, developmental changes between the ages of 6 to 12 months in their disengagement of visual attention may cause longer attention shifts that has been shown to predict later ASD diagnoses (Ibanez et al., 2008). Infants with fragile X syndrome (FXS) and infants with an older sibling diagnosed with ASD (ASIBs) are at high risk for developing ASD; thus, investigating early deficits in visual attention can help characterize the infant phenotype of FXS and ASIBs and serve as a potential prognostic indicator of ASD risk. 

Objectives: Characterize the relationship of initial latency to disengage from a non-social stimulus and frequency in attention shifts across FXS and ASIBs at 12 months of age and the potential predictive value of these attention patterns to ASD severity at 24 months. 

Methods: Participants included infant males with FXS (n=13), ASIBs; n=26), and a typically developing group (TD; n=18) assessed at 12 and 24 months of age. A toy play epoch from the Laboratory Temperament Assessment Battery (LabTAB; Goldsmith & Rothbart, 1996) reflected duration of attention to a social stimulus (mother/examiner), initial latency to disengage attention from the toy, and frequency of visual shifts in attention to the non-social stimuli (e.g., shifts away and to the toy). Behaviors were coded offline with a kappa of > 0.80. The Mullen controlled for developmental level, and the autism severity score from the Autism Diagnostic Observation Schedule-2 at the 24-month assessment represented autistic behavior. 

Results: Developmental level was included in all models. No group differences were found in the duration of gaze to the social stimuli, F(2,53)=2.48, p=0.09, partial η²= 0.09, or in their latency to disengage from the non-social stimulus, F(2,53)=1.84, p=0.17, partial η²= 0.06. Group differences were found in attentional shifts, F(2,53)=5.88, p=0.005, partial η²= 0.20, indicating infants with FXS had less attentional shifts (M=8.52, SD=2.06) than infant ASIBs (M=16.81, SD=1.35) and TD infants (M=16.79, SD=1.56). No differences were found between ASIB and TD infants (p>0.05). Regression analyses indicated that attention to the social stimuli predicted later ASD severity, F(2,23)=3.46, p=0.04, R² = 0.16 while latency to disengage did not predict later ASD severity, F(2,23)=2.23, p=0.13, R²=0.09. A trend was evident for attentional shifts predicting ASD severity, F(2,23)=3.18, p=0.06, R²=0.15. No group effects emerged for these regressions.

Conclusions: Results suggest that infants with FXS had less attention shifts than both ASIB and TD infants. Furthermore, our results support that decreased attention to social stimuli at 12 months was associated with autism symptoms at 24-months-of-age across groups. These findings suggest that atypical visual attention may be a phenotypic feature for infants with FXS and ASIBs and highlight the potential value of visual attention as a prognostic indicator of autism risk in these populations.