Author ORCID Identifier

Date Available


Year of Publication


Degree Name

Doctor of Philosophy (PhD)

Document Type

Doctoral Dissertation


Public Health


Epidemiology and Biostatistics

First Advisor

Dr. David W. Fardo

Second Advisor

Dr. Olga A. Vsevolozhskaya


Alzheimer’s disease (AD) is an irreversible, progressive brain disorder that leads to a loss of memory and thinking skills. While tremendous progress has been made in our understanding of the genetics underlying AD, currently known genetic variants explain only approximately 30% of the heritable risk of developing AD. One hurdle to AD research is that it can only be definitively diagnosed at autopsy, making cruder, clinic-based diagnoses more common. In recent years, several brain pathologies that mimic AD’s clinical presentation have been identified including brain arteriolosclerosis, hippocampal sclerosis (HS), and, most recently, limbic-predominant age-related TDP-43 encephalopathy (LATE). It has become increasingly clear that “pure AD” is rare and mixed pathologies (i.e., having two or more concomitant pathologies) are very prevalent. In this dissertation, I will present two investigations into the genetics of mixed pathologies based on autopsy-confirmed neuropathological phenotypes along with a new statistical method with an application to the genetics of mixed pathologies. The first investigation looks at a recently identified clinical AD risk locus and finds novel associations with two AD mimics and, importantly, no associations with AD-related pathologies, which suggests that the locus may be preferentially associated with non-AD dementia. The second investigation looks at the shared genetics of two related AD mimics, HS and LATE, and replicates earlier findings while also identifying several novel functional variants for both HS and LATE. The new statistical method leverages models from the branch of statistics known as functional data analysis to create a gene-level genetic pleiotropy test statistic. An extensive simulation study found that the test statistic outperforms competing methods in small sample, modest effect size scenarios and when applied to real-world data identified a novel joint association between HS and LATE and the GRN gene. All investigations were able to leverage neuropathologically-confirmed endophenotypes to identify novel genetic associations with several AD mimics, adding to the growing body of literature on the complex genetics underling neurodegenerative disease and the statistical methods available for such studies.

Digital Object Identifier (DOI)

Funding Information

This work was supported by the National Institutes of Health [R56-AG057191, R01-AG057187, P30-AG028383, R01-AG059716, K01-AG049164] from 2016 to 2021.