Brainstem white matter microstructure is associated with hyporesponsiveness and overall sensory features in autistic children

Abstract

Background: Elevated or reduced responses to sensory stimuli, known as sensory features, are common in autistic individuals and often impact quality of life. Little is known about the neurobiological basis of sensory features in autistic children. However, the brainstem may offer critical insights as it has been associated with both basic sensory processing and core features of autism.

Methods: Diffusion-weighted imaging (DWI) and parent-report of sensory features were acquired from 133 children (61 autistic children with and 72 non-autistic children, 6-11 years-old). Leveraging novel DWI processing techniques, we investigated the relationship between sensory features and white matter microstructure properties (free-water-elimination-corrected fractional anisotropy [FA] and mean diffusivity [MD]) in precisely delineated brainstem white matter tracts. Follow-up analyses assessed relationships between microstructure and sensory response patterns/modalities and analyzed whole brain white matter using voxel-based analysis.

Results: Results revealed distinct relationships between brainstem microstructure and sensory features in autistic children compared to non-autistic children. In autistic children, more prominent sensory features were generally associated with lower MD. Further, in autistic children, sensory hyporesponsiveness and tactile responsivity were strongly associated with white matter microstructure in nearly all brainstem tracts. Follow-up voxel-based analyses confirmed that these relationships were more prominent in the brainstem/cerebellum, with additional sensory-brain findings in the autistic group in the white matter of the primary motor and somatosensory cortices, the occipital lobe, the inferior parietal lobe, and the thalamic projections.

Limitations: All participants communicated via spoken language and acclimated to the sensory environment of an MRI session, which should be considered when assessing the generalizability of this work to the whole of the autism spectrum.

Conclusions: These findings suggest unique brainstem white matter contributions to sensory features in autistic children compared to non-autistic children. The brainstem correlates of sensory features underscore the potential reflex-like nature of behavioral responses to sensory stimuli in autism and have implications for how we conceptualize and address sensory features in autistic populations.

Keywords: Autism; Brainstem; DTI; Sensory features; Voxel-based analysis; White matter.

Fig. 1

Diagnosis-dependent relationships between brainstem white matter microstructure and total sensory features. Brainstem white matter tracts that exhibit total sensory feature-by-diagnosis interaction effects for the free-water-elimination mean diffusivity (FWE-MD). Analyses account for age, sex, and average head motion and apply an FDR correction for multiple comparisons. Correlations within the autistic (red circles) and non-autistic (blue triangles) groups are shown in the A corticospinal tract (CST), B medial lemniscus (ML), C lateral lemniscus (LL), D parieto-occipito-temporo-pontine tract (POTPT), E superior cerebellar peduncle cerebellothalamic tract (SCPCT), F spinothalamic tract (STT), G inferior cerebellar peduncle vestibulocerebellar tract (ICPVC), H inferior cerebellar peduncle medulla-cerebellar tract (ICPMC), and I middle cerebellar peduncle (MCP)

Fig. 2

Correlations between hyporesponsiveness and brainstem white matter microstructure in autistic children. Brainstem white matter tracts showed significant relationships between hyporesponsiveness and microstructural properties after accounting for age, sex, and average head motion and applying an FDR correction for multiple comparisons. Significant correlations were found with free-water-elimination mean diffusivity (FWE-MD) in the A medial lemniscus (ML), B lateral lemniscus (LL), C parieto-occipito-temporo-pontine tract (POTPT), D spinothalamic tract (STT), E superior cerebellar peduncle cerebellothalamic tract (SCPCT), F superior cerebellar peduncle spinocerebellar tract (SCPSC), G inferior cerebellar peduncle medulla-cerebellar tract (ICPMC), H inferior cerebellar peduncle vestibulocerebellar tract (ICPVC), I middle cerebellar peduncle (MCP), and J FWE-FA in the MCP

Fig. 3

Correlations between tactile sensitivity and brainstem white matter microstructure in autistic children. Brainstem white matter tracts showed significant relationships between tactile sensitivity and microstructural properties after accounting for age, sex, and average head motion and applying an FDR correction for multiple comparisons. Significant correlations were found with free-water-elimination mean diffusivity (FWE-MD) in the A medial lemniscus (ML), B lateral lemniscus (LL), C parieto-occipito-temporo-pontine tract (POTPT), D spinothalamic tract (STT), E superior cerebellar peduncle cerebellothalamic tract (SCPCT), F inferior cerebellar peduncle medulla-cerebellar tract (ICPMC), G inferior cerebellar peduncle vestibulocerebellar tract (ICPVC), H middle cerebellar peduncle (MCP), and I FWE-FA in the MCP

Fig. 4

Regions with distinct sensory-microstructure relationships from whole-brain voxel-based analyses with autistic and non-autistic children. Brainstem + cerebellar white matter (red) and cerebral white matter (blue) voxels indicating a significant total sensory feature by diagnostic group interaction effects represented A spatially, B as a total count, and C normalized for search space (the number of possible voxels that could be found to be significantly associated with sensory features within each area). Voxels indicating a significant total sensory feature man effects within the autistic group represented D spatially, E as a total count, and F normalized for search space