Examining brain network disruptions in African Americans and Caucasians with Alzheimer's disease

Project Summary The research so far in the diagnosis and identification of Alzheimer's disease (AD) across a variety of imaging, biomarker, and clinical modalities has been conducted studying primarily upper-class, well-educated, non- Hispanic white populations. However, Alzheimer's disease attacks individuals of all races, and African Americans over the age of 65 are more likely to be diagnosed with dementia than are any other race. African Americans have lifestyle and genetic factors that are specific to their ethnoracial group, thus making them possibly a unique AD endophenotype. Within AD, regions of the posterior default mode network tend to show lower connectivity in resting state fMRI in individuals with mild cognitive impairment (MCI) or AD. However, in our preliminary data analysis, we identified a race x diagnosis pattern: Non-Hispanic Whites exhibited lower connectivity in the MCI/AD group than in aged matched healthy controls, but African Americans with MCI and AD did not. Furthermore, we identified that only within African Americans did cerebellar, hippocampus, and putamen volumes negatively relate to total tau load. This finding leads us to hypothesize that, in African Americans, other regions of the brain may be more susceptible to disease pathology than regions of the DMN. We therefore hypothesize that along with less robust changes in CSF t-Tau, AD is associated with relatively spared DMN connectivity, but greater cerebellar, hippocampus and putamen atrophy, in African Americans with dementia. We propose to extend and replicate these results in two additional independent samples (already collected), from Emory and from the ADNI-GO/2 and ADNI-3 datasets. Using imaging, biomarker, and clinical data collected at Emory University, we will be analyzing the resting state data of approximately 150 individuals that includes African Americans and non-Hispanic whites, some of which are healthy controls or individuals with dementia. In Aim 1, we will determine in two additional cohorts (Emory-2 and ADNI-GO/2) whether race-associated changes in DMN connectivity can be replicated in older African Americans with normal cognition and MCI/AD. We will first compare functional connectivity values to diagnosis and include an interaction term of race X diagnosis in our model to determine the extent to which race modifies the relationship between functional connectivity and diagnosis. In Aim 2, we will determine whether race- associated changes in cerebellar, putamen, and hippocampal volumes' relation to CSF t-tau levels can be replicated; and if these structures show decreased functional connectivity to each other that is specific to African Americans with dementia. We will conduct a seed based analysis to investigate cerebellar, hippocampal, and putamen connectivity and its relationship to CSF biomarkers. We anticipate that, within African Americans, lower functional connectivity between the cerebellum and the putamen will be associated with higher levels of tau. This study will better characterize the African American AD endophenotype, and perhaps provide novel networks of disease pathology, specific to race.