Thyroid Function and Risk of Atrial Fibrillation and Kidney Disease - a Mendelian Randomization Study

Abstract

Overview of the thesis projects Often, the observational epidemiological study design has the limitation of being confounded, prone to selection bias, and prone to reverse causation. The Mendelian randomization (MR) design may overcome these problems as it mimicks a randomized double-blind clinical trial design. The MR design in human genetics thus relies on the random assortment of alleles at conception as it ensures random distribution of confounding factors and thereby this approach circumvents reverse causation. In the MR design, one or several genetic polymorphisms are used as proxies for a lifelong exposure to an intermediate trait. In previous observational studies, hypothyroidism and increased thyroid stimulating hormone within the reference range have been associated with reduced kidney function. Furthermore, in previous observational studies, hyperthyroidism and increased free thyroxine (fT4) have been associated with increased risk of atrial fibrillation. However, whether these observations are causal are unknown. Therefore, in this thesis, we use a MR design to assess causality for these observations. We use the Women’s Genome Health Study (WGHS) and data from two consortia, the Atrial Fibrillation Genetics Consortium (AFGen) and the Chronic Kidney Disease Genetics Consortium (CKGen). We use information from previous GWAS studies on thyroid function, kidney disease, and atrial fibrillation.