We conducted a two-sample Mendelian randomization analysis using summary data on SNPs from GWAS. We identified SNPs to proxy the protein targets of antihypertensive drugs on the basis that they mimicked the action of that drug on the target. For example, angiotensin-converting enzyme inhibitors work by inhibiting the enzyme angiotensin-converting enzyme. We therefore selected SNPs in the angiotensin-converting enzyme gene to use as a genetic proxy for the protein targets of this drug class. Effect sizes for these SNPs were then extracted from a GWAS of systolic blood pressure to estimate the instrument–exposure association.17 The instrument–outcome association was estimated using the effect sizes for these same SNPs from a GWAS of Alzheimer’s disease.18 (link) All data used were publicly available and mostly obtained from European ancestry populations [the exception being the Genotype-Tissue Expression (GTEx) project data—see below].
Full text: Click here