Age Related Differences in Hippocampal Anatomy and Correlates with Episodic Memory and Executive Function in Autism Spectrum Disorder

Background: Over the last 30 years, a rise in autism spectrum disorder (ASD) diagnoses has led to a large group of aging individuals with ASD. Little is known about how aging will affect these individuals’ neuroanatomy, compared to the neurotypical (NT) population. Recently, our group found that older adults with ASD have reduced executive functioning and smaller hippocampi compared to NT adults (Braden et al., 2017).

Objectives: This study sought to expand on previous findings from our group by segmenting the hippocampus into 12 unique subfields and determine the difference in relationships between the volumes of the individual subfields and age in ASD compared to NT adults. In addition, this study tested whether relationships between hippocampal subfield volumes and measures of executive functioning and episodic memory existed and if they were different in ASD compared to NT adults.

Methods: This project used Freesurfer 6.0 to perform a cross-sectional analysis of 12 hippocampal subfield volumes in adults with ASD (n = 53) and NT (n =39) from structural MRI scans. Using ANOVA with discrete age cohorts (young-adult: ages 18-25 years, n=42; middle-age: ages 40-70 years, n=50). Correlations were examined between subfield volumes and measures of executive function (Wisconsin Card Sorting Task [WCST] and Tower of London [ToL]), as well as episodic memory (Auditory Verbal Learning Task [AVLT]) with a false discovery rate of p=0.05 to correct for multiple comparisons. All participants with ASD had their diagnosis confirmed with the Autism Diagnostic Observation Schedule-2.

Results: There were significant interactions in several hippocampal subfield volumes (e.g. left and right hippocampal tail, subiculum, CA1, molecular layer, dentate gyrus, CA4, and fimbria, right presubiculum, and left hippocampal amygdala transition area) such that adults with ASD showed larger age-related differences compared to NT adults. Correlations with behavioral measures showed that AVLT performance was related to many subfield volumes across all participants. There were no significant relationships between hippocampal subfield volumes and WCST or ToL performance that survived correction for multiple comparisons. Correlations between behavioral measures and hippocampal subfield volumes were not significantly different between ASD and NT participants.

Conclusions: Results indicate that adults with ASD may be at risk for accelerated age-related hippocampal volume loss; however, this warrants confirmation with a longitudinal sample (in progress). Behaviorally, our study suggests hippocampal size is more strongly related to episodic memory than executive functioning performance, and that this relationship does not differ between adults with ASD and NT adults of either cohort age.