Relationship between diastolic blood pressure and Alzheimer disease: A 2-sample Mendelian randomization study based on an automated screening method for exposure factors and empirical validation

This study investigates the causal relationship between diastolic blood pressure (DBP) and Alzheimer disease (AD) using 2-sample Mendelian randomization (MR) analysis with publicly available genome-wide association study (GWAS) data and empirical validation. Automated screening identified DBP as a factor related to AD. Two-sample MR analyses were conducted using inverse variance weighting (IVW), MR-Egger regression, weighted median, simple mode, and weighted mode methods. Tests for pleiotropy, heterogeneity, and stability of genetic variants were performed. Data included DBP from Europeans (n = 422,713) and AD cases (n = 753) and controls (n = 736) of European ancestry. An empirical sample of 125 patients with cardiovascular disease in Chinese was used to verify the possibility of the above relationship by restricted cubic spline method. Thirty-seven genome-wide significant nucleotide polymorphisms from DBP GWAS were used as instrumental variables. IVW showed a causal relationship between DBP and AD (β = -1.594, SE = 0.580, P = .006). MR-Egger regression indicated minimal pleiotropy (intercept = 0.037; P = .736) and confirmed the causal relationship (β = -3.110, SE = 1.376, P = .030). The weighted median method also supported this relationship (β = -1.868, SE = 0.856, P = .029). The simple and weighted mode methods did not find significant relationships (P > .1). No heterogeneity or asymmetry was detected by Cochran Q test or the funnel plot, and the leave-one-out method showed stable results. The validation results indicate that the aforementioned findings are stable within the normal range of DBP. MR analysis supports a potential causal link between higher DBP and a lower risk of AD. The screening exposure method used could enhance MR study efficiency.