Comprehensive Gene-Environment Interactions on Alzheimer's Disease

ABSTRACT Both genetic and environmental factors are responsible for the development and progression of late-onset Alzheimer’s disease (LOAD). Although prior studies addressed G×E interactions, there are two main issues that need to be accounted for. First, many studies examined a one-gene and one-exposure model or a genome-wide loci and a one-exposure model. Second, heritability is not constant; over an individual’s lifetime genetic and environmental effects may have varying effects on LOAD risk. Beyond 1×1 interactions, constructing comprehensive scores for both genetic and environmental factors to examine G×E interactions in LOAD during midlife and later life would provide invaluable insights into vulnerability and resilience for LOAD. Polygenic risk scores (PRSs) are well-established in LOAD studies and feature an estimate of an individual’s genetic liability to LOAD by aggregating genetic effects of single-nucleotide variants (SNVs), thus serving as a comprehensive score for genetic risks [19-21]. By contrast, we currently lack optimal criteria to create analogous environmental risk scores (ERSs). To close the key knowledge gap in ERS construction, we will apply a novel approach, item function theory (IRT)-based multidimensional generalized partial credit model (GPCM) to construct midlife and later-life ERSs. We will test these hypotheses: 1) individuals with the same or a lower PRS will face an increased risk for LOAD when modified by a higher ERS (vulnerability), whereas those with the same or a higher PRS will experience a reduced risk when influenced by a lower ERS (resilience); 2) effects of PRS on LOAD differ between midlife and later-life ERSs across ethnoracial groups. To investigate these research hypotheses, we will execute the following Specific Aims: Specific Aim 1. Construct clinically diagnosed LOAD PRS. We will utilize the Alzheimer''s Disease Genetics Consortium (ADGC) genetic data and published genome-wide association studies (GWAS) summary statistics to construct optimal LOAD PRS by applying standard and penalized approaches. Specific Aim 2. Construct midlife and later-life ERSs. We will construct midlife and later-life ERSs by ethnicity/race based on various environmental factors in the National Alzheimer’s Coordinating Center (NACC) Uniform Data Set (UDS) data by employing factor analyses (for dimensional identification) and the multidimensional GPCM approach. Specific Aim 3. Examine G×E interactions on AD phenotypes. We will test the interactive effects between clinically diagnosed LOAD PRS and ERSs (midlife, later-life, and change between them) on AD phenotypes by ethnicity/race, including longitudinal cognitive changes using the NACC UDS data, and autopsy-confirmed neuropathologic changes, including amyloid-β (Aβ)/plaque, tau/tangle, α-synuclein/Lewy body, and transactive response DNA binding protein of 43 kDa (TDP-43) pathologies utilizing the NACC neuropathology (NP) data.