Longitudinal Assessment of Low Luminance Questionnaire Scores in Early and Intermediate AMD.

Autism spectrum disorder (ASD) is a neurodevelopmental disorder characterized by deficits in social skills and the presence of restricted and repetitive behaviors or interests (). The great heterogeneity in its clinical manifestations and behavioral profiles poses major challenges for early diagnosis and effective intervention. Unveiling the neurobiological basis underlying this heterogeneity and identifying the potential brain markers have therefore been a key focus of contemporary neuroscience (). As ASD follows a developmental trajectory (; ), childhood represents a crucial window for characterizing atypical brain development and identifying early biological signatures of the disorder. The investigation of alterations in brain structure has been paramount, as structural connections reveal the physical architecture that supports and constrains brain functions (; ). White matter, in particular, plays a crucial role in shaping functional synchronization and supporting early cognitive development (). For example, greater microstructural integrity of white matter tracts during the first year of life was associated with better cognitive and language outcomes at ages 1 and 2 (). Previous studies have consistently found that individuals with ASD exhibit altered brain structure compared to healthy controls (; ; ). However, these studies reported primarily on localized changes, such as the integrity of specific tracts or abnormalities in regional gray matter volumes, rather than examining the system-level organization. The brain, however, functions as an integrated network rather than a collection of isolated regions. Emerging evidence suggests that atypical structural patterns in ASD span multiple brain systems (M. ; ), underscoring the importance of using a network-based approach to explore the neurobiological basis of ASD. Modelling the brain as a large-scale network enables quantification of topological properties that might be more sensitive to the widespread alterations in ASD. Despite growing interest in network-based approaches, previous studies have yielded inconsistent and sometimes contradictory findings. Some reported decreased network efficiency in ASD (; ), whereas others observed increased efficiency (; S.-J. ; ), while several found no significant group differences (; ). These inconsistencies may result from the substantial clinical heterogeneity within the autism spectrum, where different ASD subtypes may reflect unique brain systems and behavioral phenotypes (). Treating ASD as a single, homogeneous category likely obscures meaningful neurobiological variation, contributing to these inconsistent results. Therefore, stratifying ASD into meaningful subgroups and using graph-theoretic methods may offer deeper insights into the etiology of the disorder ().