Figure 1
Schematic diagram of Mendelian randomization (MR) experiment of exposure upon outcome of interest. Known genetic instruments (SNPs) for exposure were used as instrumental variables to assess the causal effect of exposure upon outcomes. Β1exposure is the estimated coefficient from the regression of exposure on the genetic variant(s) using the exposure GWAS. Β2(Observed) is the observed coefficient of the relationship between the exposure and outcome of interest. Β3outcome is the regression coefficient of outcome on the corresponding genetic variant using the outcome GWAS. MR estimate is the wald ratio of Β3outcome/Β1exposure. Significant (P < 1 × 10−8) and near-independent BMI SNPs (based on an approximate conditional and joint multiple-SNP (COJO) analysis that takes into account LD (linkage disequilibrium) between SNPs at a given locus) from BMI GWAS were used as instrumental variables to assess the causal effect of BMI upon AD while significant (P < 5 × 10−8) and independent AD SNPs (at least 4 MB(mega base pairs) apart) from AD GWAS were used as instrumental variables to assess the causal effect of AD upon BMI.
