Older Adults with ASD: Executive Functioning Deficits, Functional and Structural Connectivity Differences, and Accelerated Cortical Thickness Atrophy

Background:  The adult population of autism spectrum disorder (ASD) is rapidly growing and continuing to get older, yet there are few studies of aging in ASD to inform appropriate care and biological targets for intervention. ASD and normal aging have similarities; executive functioning deficits are commonly associated with ASD and declines are also observed in normal aging. This decline is often attributed to greater rates of age-related atrophy in the frontal lobe, which largely subserves executive functions. There is also an anterior-to-posterior gradient in brain pathology related to ASD, which implicates the frontal lobe as the most affected region. Additionally, reduced integrity of white matter fibers supporting the frontal lobe are evident and functional connectivity disturbances are largely characterized as anterior-posterior hypoconnectivity in ASD and also observed in normal aging.

Objectives:  Given the striking cognitive and neuroanatomic parallels in ASD and aging, it is hypothesized that ASD exacerbates age-related executive functioning deficits, frontal lobe integrity, and functional connectivity differences, compared with young-adult and aged typically developing (TD) controls. The study objective is to address the current gap in knowledge of the aging ASD cohort through available cross-sectional structural imaging data from the Autism Brain Imaging Data Exchange (ABIDE) and pilot cognitive and imaging analysis of middle-age ASD and TD cohorts. 

Methods:  From the open-access ABIDE data repository, 2.8% of the sample was middle-aged (39-58; 15 ASD, 15 TD). There were no significant group differences in IQ or age between ASD and TD and we matched middle-age participants with young-adults (18-25) based on gender, ASD diagnosis (autism vs. Asperger’s vs. PDD-NOS), and IQ. We examined cortical thickness in FreeSurfer (http://surfer.nmr.mgh.harvard.edu/) for an interaction between age and diagnosis. We also examined  differences in 9 middle-age ASD and 9 TD control males on executive functioning via the Wisconsin Card Sorting Task (WCST), frontal lobe structural connectivity via fractional anisotropy (FA), and anterior-posterior resting-state functional connectivity (rsFC).

Results:  For cortical thickness using the ABIDE data, we observed an anterior-to-posterior gradient of interaction, such that effects were most pronounced in the frontal and temporal lobes (Figure 1a). Using automated segmentation we extracted regional averages and detected a pattern of thicker anterior cortices in young-adult ASD versus young-adult TD, but thinner cortices in middle-age ASD, as compared with TD, suggesting an exacerbation of age-related cortical thickness loss in ASD (Figure 1b). Within our sample of middle-age men, the ASD group made more perseverative errors on the WCST (Figure 2a), had reduced FA in bilateral corona radiate (Figure 2b), and decreased rsFC between medial prefrontal cortex and posterior cingulate cortex, than the TD group (Figure 2c). 

Conclusions:  Taken together, preliminary evidence from the ABIDE dataset suggests an anterior to posterior exacerbation of age-related cortical thickness loss in ASD; and preliminary evidence from our sample confirms the presence of robust executive function deficits, decreased integrity of anterior white matter tracts, and anterior to posterior resting-state hypoconnectivity in an aging ASD cohort, which warrants further cross-sectional and longitudinal investigation of age-related changes.