Back

Altered sensorimotor-association axis patterning of global functional connectivity in an autism subtype with low levels of language, intellectual, and adaptive functioning

Severino, I.; Mandelli, V.; Bertelsen, N.; Lombardo, M. V.

2025-11-02 psychiatry and clinical psychology
10.1101/2025.10.30.25339153 medRxiv
Show abstract

In early development autism can be stratified into subtypes differentiated by non-core language, intellectual, motor, and adaptive functioning features. In toddlerhood, these subtypes show different genomic patterning effects on brain structure and function that follow early primary axes of neurodevelopmental organization. This leads to the hypothesis that early cortical patterning differences may continue to be evident between subtypes in later development within hierarchical organization along the sensorimotor-association (S-A) axis. To test this hypothesis, we first demonstrate that unsupervised data-driven techniques can detect 2 autism subtypes (high versus low) with very high accuracy (92%) in late childhood to adulthood based on language, intellectual, and adaptive functioning features. In resting state fMRI data, we find that global functional connectivity is differentially patterned in the subtypes along the S-A axis. Autism with lower levels of language, intellectual, and adaptive functioning have hypo-connected sensorimotor and hyper-connected association areas relative to a non-autistic comparison group. These findings suggest that cortical patterning of global functional connectivity along the S-A axis differentiates an autism subtype with low levels language, intellectual, and adaptive functioning skills.

Published in Biological Psychiatry: Cognitive Neuroscience and Neuroimaging (predicted rank #3) · training set

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.