Analysis of pleiotropic genetic effects on cognitive decline and systemic inflammation in the Women’s Health Initiative Memory Study

Abstract

Listen

Translate article

SNPs associated with C-reactive protein (CRP) and plasma lipids have been investigated for polygenic overlap with Alzheimer's disease (AD) risk SNPs. Previously, we reported pleiotropic effects between SNPs associated with cognitive impairment (CI) and SNPs associated with systemic inflammation as measured by CRP and plasma lipids in the Health and Retirement Study (HRS). We sought to replicate our results in the Women's Health Initiative Memory Study (WHIMS). Analysis of SNP pleiotropic effects was completed using data from 6,078 non-Hispanic White women (aged 71.4 [SD 7.4] at baseline) for whom genetic, cognitive, and/or biomarker data were available. We identified CI using WHIMS adjudicated case status (normal vs. mild cognitive impairment or probable dementia). Secondary phenotypes included baseline CRP and plasma lipid levels (HDL, LDL, and total cholesterol [TC]). Genome-wide association analyses were conducted for all phenotypes. A conditional false discovery rate framework defined genetic pleiotropy. Genetic pleiotropy was identified if SNPs were associated with CI conditional on association with secondary phenotypes (CRP, HDL, LDL, TC) at an FDR < 0.05. Functional genomic bioinformatics analysis and comparison with prior findings were conducted. Genetic pleiotropy was observed for CI conditional on association with the secondary phenotypes of plasma CRP, HDL, LDL, and TC. These results replicate associations for SNPs in the APOE, APOC1, PVRL2/NECTIN2 loci reported in the prior HRS study for CRP, LDL, TC. The odds ratios for the association of these SNPs with CI ranged from 0.5 to 4.0, consistent with moderate effect sizes for the APOE ε2 or ε4 alleles. For WHIMS, FDR significant results were observed in both directions for several of the traits (CRP, lipids), further supporting evidence of genetic pleiotropy. Moreover, we observed associations for SNPs in the APOE, APOC1, PVRL2/NECTIN2 loci for CI conditional on HDL. We identified polygenic overlap between CI and CRP, LDL, HDL, and TC phenotypes in WHIMS, largely replicating prior results in the HRS. The variants and associated genes identified are involved in pathological processes including metabolic, cardiovascular, and immune response and are potentially important for AD risk prediction and development of therapeutic approaches based on anti-inflammatory mechanisms.