Abnormalities of Local Gyrification in Autism Spectrum Disorders after Age 40 Years

Background: Although prenatal or early postnatal in onset, autism spectrum disorders (ASDs) are associated with behavioral problems throughout the lifespan. Neurodevelopmentally, evidence shows early cerebral overgrowth in the first years of life, reflected in cortical thickness (CT), surface area (SA), and the local gyrification index (lGI). However, abnormalities across later childhood and adolescence are less well understood, and little is known about adults with ASDs past the age of 40 years, including possible accelerated decline. A few studies extending into early to middle adulthood suggested increased rates of cortical thinning in ASDs in frontal, parietal, and temporal regions. One study reported increased frontal and parietal lGI in participants with ASDs aged 18-43 years, while another detected no differences in gray matter characteristics in individuals aged 30-75 years.

Objectives: The current investigated cortical morphology in adults with ASDs ages ≥40 years, in a comparison with age-matched typical controls (TC).

Methods: T1 weighted MRI sequences (0.8mm isotropic) were collected for 50 participants aged 40–61 years (27 ASD, 23 TC). Following quality assurance, 20 ASD and 21 TC, matched on age, were compared. lGI, CT, and SA were measured using FreeSurfer v.5.3.0. Statistical analyses employed a general linear model including age, non-verbal IQ, and total brain volume as covariates. Vertex-wise correlations between lGI and Social Responsiveness Scale (SRS) scores were calculated. All results were corrected for multiple comparisons using Monte Carlo null-z simulations.

Results: For lGI, significant main effects of group (ASD<TC) were observed bilaterally in insular and anterior cingulate (ACC) clusters, along with left postcentral and middle frontal, and right orbitofrontal and supramarginal clusters. lGI was also shown to decline with age in combined groups in bilateral precentral and right supramarginal clusters. No significant group, age, or group by age interaction effects were observed for CT or SA. lGI showed a significant correlation with SRS total scores in a right caudal ACC cluster (r=.68, p<.001) in the TC group only.

Conclusions: Reduced lGI may suggest accelerated tissue loss in ASDs, possibly consistent with some studies reporting increased cortical thinning in early adulthood. The finding of lGI differences in the absence of CT and SA effects indicates that lGI may be more sensitive to abnormalities of cortical macrostructure in ASDs. Clusters in the insula overlapped with a previous study of children and adolescents, in which lGI was increased in two independent samples of participants with ASDs in regions surrounding earlier developing sulci, and perisylvian regions in particular. This overlap may indicate that regions most affected by early over-gyrification are also more susceptible to subsequent decline in the disorder. Group differences in ACC are of interest given multiple post-mortem reports of its altered cellular density in ASDs and the importance of ACC in social and emotional function. Correlations between lGI and SRS scores in TC in the same region add additional support to the functional relevance of this morphometric measure.