Abstract
BACKGROUND: Structural alterations in subcortical brain regions-including the amygdala, hippocampus, basal ganglia, and cerebral ventricles-have been linked to various clinical features of autism spectrum disorder (ASD). However, volumetric features among these regions in autistic individuals across the lifespan remain poorly understood. This cross-sectional study aimed to investigate age-associated volumetric deviations in these clinically implicated subcortical regions of autistic individuals and neurotypical controls, and to examine the structural interrelationships within each group. METHODS: We examined multi-site T1-weighted MRI data from 119 autistic and 122 neurotypical participants aged 7-73 years. Volumetric data for the amygdala, hippocampus, basal ganglia, and cerebral ventricles were harmonized across sites using the ComBat algorithm. Following this, volumetric composite indices (principal component scores) were extracted for each region using principal component analysis. These scores represent dominant volumetric patterns of each subcortical region, with higher values reflecting greater volume. These composite scores were then compared between groups and with increasing age. RESULTS: Autistic participants exhibited greater amygdala volume in early life, followed by more pronounced age-associated reductions in adulthood compared to neurotypical controls. A similar trend was observed for the hippocampus, with early volumetric enlargement giving way to steeper declines in later years. In contrast, the autistic group consistently trended towards larger basal ganglia across the lifespan. Additionally, autistic participants showed accelerated enlargement in the cerebral ventricles with increasing age. Both groups exhibited patterns of inverse volumetric associations between the cerebral ventricles and surrounding subcortical regions in later adulthood; however, these relationships were more pronounced and widely distributed in the autistic group. LIMITATIONS: The cross-sectional design of this study limited us from capturing intra-individual differences at baseline and quantifying the lifespan trajectories of each participant. Site-related sampling differences may have introduced cohort bias in the results. CONCLUSIONS: Autistic participants and neurotypical controls exhibit distinct, age-related volumetric patterns across key subcortical brain regions. Enlargement of the cerebral ventricles and their inverse structural relationships with neighboring structures in later life may indicate atrophic processes beginning in middle adulthood in ASD. These findings highlight the need to further investigate mechanisms of atypical brain aging in ASD and consider these subcortical brain regions as potential biomarkers of neurodegeneration and intervention targets across the adult lifespan.