Principal Investigator: Dr Sarah Pendergrass
Geisinger Clinic, PA, USATags: 49945, genomics, GWAS, Mendelian randomisation, obesity, PheWAS
Understanding the biggest risk factors for disease, and the biological reasons disease occurs, are important to improving how we treat disease as well as target intervention strategies. We know that weight is an important contributor to risk of other conditions, such as the risk of cardiovascular disease. However, there is still to more to be understood about the relationship between weight and other health outcomes. For example, we know that high weight contributes to a higher risk of cardiovascular disease. There are many other conditions that are known to more likely co-occur with obesity. Understanding more about the complex relationship between weight and detrimental health outcomes will highlight conditions where intervening on weight will have the most impact on other conditions.
For this study we will use data from the UKBiobank as well as data from the Geisinger MyCode Community Health Initiative. Both resources have considerable information on diseases, health outcomes, and other measures. Geisinger is the largest health care provider in central Pennsylvania, covers millions of patients, and has an electronic health record (EHR) system that was started in 1996. Geisinger also has existing genetic resources for ~130,000 individuals. By using data from both of these resources we will be able to compare and contrast results between the two studies. There are similarities in ancestry for both datasets, and yet different environmental exposures.
We will evaluate the relationship between BMI and a wide range of outcomes and relevant lab and medical measures using statistical analyses in both the UKBiobank and Geisinger. Then we will characterize the relationship between genetic variation across the genome and BMI. We will also evaluate the association between genetic variation across the genome and the outcomes and phenotypes most statistically associated with BMI. We will use a statistical approaches to then infer causal relationships between BMI and the most statistically associated traits and conditions we identified for BMI. This statistical approach provides information on the shared co-occurrence of traits based on shared genetics, clarifying the direct relationship between two co-occuring traits.