Relation of genomic variants for Alzheimer disease dementia to common neuropathologies

Abstract

Objective: To investigate the associations of previously reported Alzheimer disease (AD) dementia genomic variants with common neuropathologies.

Methods: This is a postmortem study including 1,017 autopsied participants from 2 clinicopathologic cohorts. Analyses focused on 22 genomic variants associated with AD dementia in large-scale case-control genome-wide association study (GWAS) meta-analyses. The neuropathologic traits of interest were a pathologic diagnosis of AD according to NIA-Reagan criteria, macroscopic and microscopic infarcts, Lewy bodies (LB), and hippocampal sclerosis. For each variant, multiple logistic regression was used to investigate its association with neuropathologic traits, adjusting for age, sex, and subpopulation structure. We also conducted power analyses to estimate the sample sizes required to detect genome-wide significance (p < 5 × 10(-8)) for pathologic AD for all variants.

Results: APOE ε4 allele was associated with greater odds of pathologic AD (odds ratio [OR] 3.82, 95% confidence interval [CI] 2.67-5.46, p = 1.9 × 10(-13)), while ε2 allele was associated with lower odds of pathologic AD (OR 0.42, 95% CI 0.30-0.61, p = 3.1 × 10(-6)). Four additional genomic variants including rs6656401 (CR1), rs1476679 (ZCWPW1), rs35349669 (INPP5D), and rs17125944 (FERMT2) had p values less than 0.05. Remarkably, half of the previously reported AD dementia variants are not likely to be detected for association with pathologic AD with a sample size in excess of the largest GWAS meta-analyses of AD dementia.

Conclusions: Many recently discovered genomic variants for AD dementia are not associated with the pathology of AD. Some genomic variants for AD dementia appear to be associated with other common neuropathologies.

Figure 1. Model illustrates the association of risk factors with neuropathology and AD dementia

Conceptual model shows that a risk factor previously related to AD dementia may be associated with pathologic AD, with other coexisting neuropathologies, or not with any known neuropathology (association with reserve or resilience markers). AD = Alzheimer disease; MCI = mild cognitive impairment.

Figure 2. Association of genomic variants with non-AD pathologies

Plot shows the association of genomic variants with non-AD pathologies (horizontal line marks p < 0.05). Multiple logistic regression analysis adjusted for age at death, sex, and the first 3 principal components from EIGENSTRAT. The y-axis shows the −log10 of the p values, such that the higher the value, the more significant the association. AD = Alzheimer disease.

Figure 3. Sample size calculation for association of genomic variants with pathologic AD (truncated at n = 75,000)

Plot shows sample size estimation for association of genomic variants with pathologic AD (truncated at n = 75,000). Sample size required to reach genome-wide significance (p < 5 × 10⁻⁸) with 80% power. The sample size calculation depends on the case-to-control ratio, frequency of the variant (with respect to the coded allele), baseline disease risk, as well as the effect size (estimated from logistic regression models adjusted for age at death, sex, and the first 3 principal components from EIGENSTRAT). The case-to-control ratio was estimated by the ratio of total numbers of participants with pathologic AD diagnosis (considered as cases) and those without (considered as controls). The baseline risk of pathologic AD was the estimated probability of AD for a typical woman with mean age 89 years and absence of the coded allele. Colors represent different chromosomes. AD = Alzheimer disease.