Development of Visual Attention in Infants with Increased ASD Risk: A Longitudinal Assessment

Background:  

Infants who have an older sibling with ASD are at greater risk of receiving an autism diagnosis than the general population (Ozonoff et al., 2011). By identifying early risk markers, diagnosis and intervention can occur earlier and may increase effectiveness. It has been suggested that atypical visual attention may be an early risk marker, yet the evolution of visual attention in at-risk infants over time is not well known (Elsabbagh et al., 2007). Tracking the differences in attention performance between high-risk and low-risk infants longitudinally may help identify the age at which visual attention is most vulnerable.

Objectives:  

To examine visual attention performance in infants with (high-risk) and without (low-risk) siblings with ASD at 6, 12, and 18 months.

Methods:  

Participants were 24 high-risk siblings with an older sibling with ASD and 37 low-risk siblings with no family history of ASD. Assessments occurred at 6, 12, and 18 months. Both groups participated in the “gap-overlap” task, where reaction time to disengage from a central to peripheral stimulus was measured. There were three trial types: baseline (central stimulus disappears as peripheral stimulus appears), overlap (central stimulus remains on screen with peripheral stimulus), and gap (central stimulus disappears before the peripheral stimulus appears). The difference in overlap and baseline trials is believed to reflect efficiency in attention disengagement. Differences between gap and baseline trials are thought to indicate recognition of visual facilitation.

The task was repeated twice, with different stimulus sets, at each age point. At Visit 1, the task was administered at the end of the visit (approximately 1.5 hours after arrival) and at Visit 2 the task was administered before the visit. All data at 6 and 12 months have been collected; collection at 18 months is ongoing.

Results:  

The reaction time for the overlap trial was significantly slower than the baseline and gap trials for both high-risk and low-risk groups. Both groups also showed a significantly faster reaction time in the gap trial than baseline or overlap trials. Preliminary analyses showed that 6-month-old high-risk infants, when performing the task at the beginning of the visit, showed similar disengagement trends to the control infants. When the task was performed after participating in other tasks, the high-risk group showed significantly slower disengagement than the low-risk group, replicating previous reports (Elsabbagh et al., 2009).  

Conclusions:

The significant effect of trial type on 6-month-old infants’ reaction time replicates previous findings. At 6 months, the high-risk group demonstrates greater variability in performance depending on when the measurement was collected. Further analysis of 12 month and 18 month data will address age-related effects.