Infants at High Risk for ASD Show Altered Functional Connectivity in Auditory Networks at 9 Months of Age

Background: Altered functional connectivity is characteristic of developmental disorders involving language impairments such as autism spectrum disorder (ASD). Children with ASD show altered connectivity in several resting state networks compared to neurotypical controls (Uddin et al., 2013, Hernandez et al., 2014), suggesting that functional connectivity may provide a reliable biomarker of the disorder. In particular, connectivity with the right anterior insula, the hub of the salience network, is altered in ASD. In addition to its role in modulating attention to salient elements of one’s environment, the insula has also been shown to be specialized for sensory and auditory-related functions as an integration hub for auditory stimuli (Bamiou et al., 2003; Oh et al., 2014). As such, altered functional connectivity between auditory regions and the insula early in life may predict future language outcome and ASD diagnosis.

Objectives: Here we use resting-state fMRI (rsfMRI) to characterize the developing auditory network and its connectivity with the right anterior insula in 9-month-old infants at high (HR) and low risk (LR) for ASD. Early differences in functional connectivity may predict altered developmental trajectories prior to overt language delays and the onset of ASD symptomatology.

Methods: During natural sleep, 8-minute rsfMRI scans were collected in 9-month-old infants. Preprocessing was performed using FSL. To correct for effects of head motion, an advanced ICA-based strategy for automatic removal of motion artifacts (ICA-AROMA; Pruim et al., 2015) was used. Auditory network connectivity was examined using left and right hemisphere Heschl’s gyrus (HG) and superior temporal gyrus (STG) seeds from an infant brain template (Shi et al., 2011). Time-series extracted from processed residuals in standard space were correlated with every other voxel to generate functional connectivity maps. We examined group differences for functional networks associated with each seed separately. Next, we investigated connectivity between primary auditory cortex (left and right HG) and right anterior insula, the hub of the salience network. Lastly, differences in auditory network connectivity were related to behavioral indices of language development and ASD symptom severity collected at 12 and 18 months of age.

Results: Whole-brain correlation maps generated from left and right HG and STG seeds demonstrated that the auditory network was detectable in all infants. Between-group comparisons yielded significant differences in functional connectivity with HR infants showing hyperconnectivity between left HG and supplementary motor area. By contrast, LR infants showed stronger intrahemispheric connectivity between right STG and right temporo-parietal areas when compared to the HR group. Furthermore, HR infants showed hyperconnectivity between both left and right HG with right anterior insula. Auditory network connectivity and auditory-insular hyperconnectivity were positively associated with better language outcome at both 12 and 18 months, and less severe ASD symptomatology at 18 months.

Conclusions: At 9 months of age, the auditory network is hyperconnected with various cortical and subcortical regions in infants at risk for developing ASD. Early evidence of increased connectivity between these regions may reflect a compensatory mechanism in ASD risk.