Characterisation of Associations between Social and Non-Social Attention and Later ASD Symptoms in Infants with Tuberous Sclerosis Complex: An Eye-Tracking Study

Background: Tuberous sclerosis complex (TSC) is a single-gene disorder associated with variable developmental outcomes, with up to 60% receiving a diagnosis of autism spectrum disorder (ASD). The specific mechanisms through which individuals will go on to develop ASD are yet to be clearly defined. An increasing proportion of TSC cases are identified prenatally, providing a unique opportunity to study risk pathways through infancy. Social and non-social attention have been implicated in infants at familial risk for ASD. The degree of homology of association between these attentional parameters and risk for ASD, and the role of early-onset epilepsy, in TSC is unknown.

Objectives: To characterise differences in social and non-social attention in infants with TSC and typically developing infants and explore associations with epilepsy and emerging ASD symptoms.

Methods: The Early Development in Tuberous Sclerosis (EDiTS) Study is an ongoing prospective longitudinal study of infants with TSC (total n=25) and age-matched typically developing infants (total n=25) enrolled from birth to 24 months of age. Infants undergo developmental assessments (e.g. Mullen Scales of Early Learning) at each visit. We use portable eye-tracking technology in the home environment to measure social and non-social attention using a range of tasks. In this study, we focused on two paradigms that have been used in several previous infant studies: a face ‘pop-out’ task, to measure proportion of time looking at faces, and a gap/overlap task, to measure the disengagement effect (response time in baseline versus overlap).

Results: Data is currently available for 10 infants with TSC and 15 typical infants at 5, 8, 10 and 14 months old, with analysis ongoing. Infants with TSC show a slower gain in skills, particularly in the non-verbal domain from 5 to 10 months of age (p=.007) associated with younger age at seizure onset (rho=.61, p=.04). At 5 months, infants with TSC display increased peak looking time to faces compared to typical infants (2.42 versus 4.17, p=.03), associated with younger age at seizure onset (rho=-.47, p=.08), but not with ASD symptom scores at 14 months (Q-CHAT). Conversely, infants with TSC show longer disengagement times on the gap/overlap task from 10 months of age compared to typical infants (254ms versus 177ms, p<.05), which was associated with increased ASD symptoms at 14 months (rho=.60, p=.05).

Conclusions: These preliminary findings indicate atypical social visual attention from 5 months combined with reduced flexibility in control of non-social visual attention from 10 months in infants with TSC, suggesting early developmental changes that partially mimic ASD risk markers in familial cases. The variable timing in the onset of social versus non-social attentional differences suggest the interaction of multiple brain systems over the first year of life that give rise to variable pathways. In TSC, variation may reflect the differential effect of epilepsy on low- and high-order cognitive domains. Identification of sensitive and objective biomarkers of risk for ASD in TSC will aid in directing more specific interventions and in testing the efficacy of novel early intervention targets.