The aim of this study to investigate the role of potential risk factors for Alzheimer's (AD) using Mendelian Randomization. Smoking, hypertension, increased body weight, dyslipidemia and type 2 diabetes have all been suggested as risk factors for AD but the results of epidemiological studies on them have been inconclusive.

In this project we are planning to use genetic variants that are robustly associated with the above exposures to test if they have a causal effect on AD through Mendelian randomization (MR). For blood pressure, BMI, lipids and type 2 diabetes the method of analysis will be a 2-sample MR in which the "reduced-form" estimate (the coefficient for the association between the IV and the outcome) and the "first-stage" estimate (the coefficient for the association between the IV and the exposure) are obtained from non-overlapping samples (Pierce and Burgess 2013). The instrumental variables for the MR analysis will consist of an allelic score calculated based on the results of independent genome-wide association studies of the above traits.

Polygenic risk scores will be calculated based on the results of published GWAS studies. These identified risk alleles will be used to calculate a polygenic score for each individual in an independent (target) sample (ALSPAC), corresponding to the mean number of score alleles (weighted by the logarithm of the odds ratio) across the set of SNPs, using PLINK. Standard linear regression models will be used to calculate coefficients for the association of the polygenic scores with the trait of interest in the ALSPAC target sample (first-stage coefficients). The reduced-form coefficients of the genetic variants/polygenic scores will be calculated in the International Genetics of Alzheimer's Project (IGAP) sample, which is independent of the target sample where the first-stage coefficients would be calculated. The IGAP sample is a collaboration of 4 groups in Europe and USA with approximately 60,000 individuals (either with AD or controls). Genetic data are available on all of them. We have obtained permission to perform this analysis in the IGAP sample.