A disseratation defense from Sarah Kolibash Royse, Public Health/Epidemiology. 

Committee: 

Caterina Rosano (dissertation chair, co-advisor), Epidemiology
Beth E. Snitz, dissertation co-advisor), Neurology
Marnie Bertolet, Epidemiology and Biostatistics
Ann D. Cohen, Psychiatry
Ashley V. Hill, Epidemiology
Joseph M. Zmuda, Epidemiology and Human Genetics

Abstract: 

African American (AA) populations are disproportionately burdened by clinical Alzheimer’s disease (AD), but do not show more evidence of pathological AD than those who are non-Hispanic white (nHW), suggesting that current pathological disease models are insufficient to explain this disparity. I used a population neuroscience approach to interrogate three potential drivers of clinical AD disparities, where each driver represented group of clinical AD risk factors: white matter hyperintensities (WMH; neuropathological risk), apolipoprotein-E genotype (APOE; genetic risk), and perceived discrimination (social/contextual risk). In the first study, I used a novel method to quantify WMH, unhealthy white matter connectivity (UWMC), and examined overall and racialized group differences in the association of UWMC in AD pathology-affected brain regions with cognition. Compared to nHW participants, those who were AA exhibited more UWMC, worse cognitive function, and similar UWMC-and-cognitive function relationships, indicating that racialized group disparities in clinical AD may be driven partially by differential burden of WMH/UWMC. The second study examined overall and racialized group differences in relationships of APOE*4 and APOE*2 with in vivo Aβ and tau. Relative to those who were nHW, AA participants were more likely to carry at least one APOE*4 or APOE*2 allele, exhibited less Aβ, and showed a weaker (non-significant) relationship between APOE*4 and Aβ; all other relationships were non-significant. Thus, APOE*4 may not drive racialized group disparities in clinical AD through risk of pathological AD; the role of APOE*2 deserves further study. Finally, I tested overall and racialized group differences in associations of mid-life perceived discrimination with late-life AD pathology and cognition. AA participants exhibited more mid-life perceived discrimination, worse late-life executive function, and a stronger relationship between mid-life perceived discrimination and late-life executive function than nHW participants; no other associations were significant. Clinical AD disparities may therefore be partially driven by differential burden and consequences of mid-life perceived discrimination in AA populations. Overall, I found that UWMC in AD pathology-affected regions may contribute to clinical AD disparities. Interactions between UWMC and AD pathology on clinical AD and whether APOE*4 and perceived discrimination and other social/contextual exposure influence UWMC through conferring vascular risks should be studied further.