The Aging Autistic Brain: Multi-modal imaging to predict accelerated memory decline - Project Summary/Abstract Middle-age adults with autism spectrum disorder (ASD) have a 2.6x higher risk for early-onset Alzheimer’s disease and other dementias than non-ASD. In children, ASD prevalence is ~2%. Since ASD is a lifelong condition, there will soon be a large population of elderly with an ASD diagnosis (~700,000 in the U.S within the next decade), with a lifetime cost of ~$3.6 million per person. In a series of two papers, we recently published accelerated longitudinal decline of memory and hippocampal volume in middle-age and older adults (MA+) with ASD, compared to matched controls. This adds to our cross-sectional findings of decreased hippocampal functional connectivity and increased free-water. Free-water is a relatively novel microstructural diffusion tensor imaging technique that is more sensitive than conventional MRI metrics for detecting neurodegeneration. Our recent paper shows baseline hippocampal free-water is a stronger correlate of memory decline in our MA+ ASD than hippocampal volume. For functional MRI (fMRI), it is notoriously difficult to efficiently elicit strong hippocampal activation. Nevertheless, altered hippocampal activity and connectivity have been found in adults with ASD. We developed a relatively brief visuospatial fMRI memory task and our preliminary data show strong single subject MA+ ASD hippocampal signal. We will add this novel task and an established verbal memory task to our ongoing longitudinal MA+ ASD study. We will link fMRI data with structural MRI measures using our novel multi-modal analysis methods to identify patterns that best predict accelerated memory decline. Our long-term goal is identifying biomarkers and intervention targets for precision medicine care of aging autistic adults. The specific aims are to: 1) characterize hippocampal and memory aging trajectories in MA+ ASD, compared to matched NT adults, and 2) identify the combination of MRI measures that best predict accelerated memory decline in MA+ ASD, compared to matched NT. Our study design will use our ongoing, well-characterized longitudinal cohort of MA+ ASD (n=62) and NT (n=61) who undergo cognitive batteries and multi-modal MRI scans every two-years (ages 40-75, mean=55.2 years; M:F ratio 1.6:1; 1-4 timepoints). Our analysis approach will be a combination of hypothesis-driven, longitudinal multi-level models with hippocampal and memory scores of interest longitudinal, and exploratory whole-brain longitudinal and multi-modal partial least squares analyses. Our expected outcome is to lay the groundwork for: 1) multi-modal MRI biomarker development that can be used to predict cognitive aging outcomes, and 2) identifying targets for neuromodulation or other interventions. This is relevant to public health because findings will advance fundamental knowledge of brain aging vulnerabilities and mechanisms in MA+ ASD. This will lay the groundwork for precision medicine advances at the intersection of neuro-developmental conditions and cognitive aging, and serve as a model for other developmental conditions (e.g. Attention-deficit/ hyperactivity disorder, Cerebral Palsy).
