Integrating genetic correlation, polygenic risk scores, and Mendelian randomization to identify modifiable risk factors for Alzheimer's disease. (20th December 2022)

• Record Type: Journal Article
• Authors: Andrews, Shea J; Fulton‐Howard, Brian; Bennett, David A; Goate, Alison
• Is Part Of: Alzheimer's & dementia. Volume 18(2022)Supplement 11
• Page Start: n/a
• Page End: n/a
• Publication Date: 2022-12-20
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.065084
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: To reduce the population prevalence of Alzheimer's disease (AD), it is critical to identify risk factors that modify AD risk. Methods of causal inference that exploit genomic information, such as genetic correlation (rg ), polygenic risk scores (PRS) and Mendelian randomization (MR), can overcome some of the limitations of observational studies such as confounding and reverse causation. Here we use rg, PRS, and MR to investigate causal associations between twenty‐two risk factors and eleven AD outcomes. Methods: Using GWAS summary statistics, rg was estimated using GNOVA and bidirectional MR causal estimates using LHC‐MR. The exposures included alcohol intake, physical activity, lipid traits, blood pressure traits, diabetes, BMI, depression, insomnia, social isolation, smoking, diet, and education. The outcomes included AD, AD age of onset, hippocampal volume, CSF Aβ42, tau, and Ptau181 levels, and amyloid, tau, cerebrovascular neuropathology. PRS for the exposures were constructed using PRSice in ADGC (n = 25, 431), ADNI (nmax = 1, 718), and ROSMAP (nmax = 1, 675), and their association with AD diagnosis, CSF biomarkers, PET imaging, MRI imaging, and neuropathology was evaluated. Results: After accounting for multiple testing, 71, 1, and 38 trait pairs were significant in the rg, PRS, MR models respectively. The only significant trait pair identified across all three methods was a protective effect of higher educational attainment on AD (rg = ‐0.161, p = 3.78e‐14; bPRS [se] = ‐0.07 [0.014], p = 8.88e‐7; bMR [se] = ‐0.21 [0.042], p = 2.43e‐7). Seventeen trait pairs had both significant genetic correlations and causal MR estimates. In particular, increased cortical surface area was positively correlated with higher educational attainment (rg = 0.261, p = 1.46e‐18), with MR indicating that genetically predicted cortical surface area was causally associated with higher education (bMR [se] = 0.27 [0.042], p = 3.21e‐10), but not vice versa (bMR [se] = ‐0.09 [0.072], p = 0.224). Conclusions: We evaluated causal relationships between risk factors and AD endophenotypes using genomic information. The protective effect of education on AD is supported by genetic correlation, PRS and MR, potentially due to increased cognitive resilience resulting from increased cortical surface area.