Cardiovascular diseases are the global leading cause of morbidity and mortality, accounting for 32% of deaths (> 17 million) annually as of 2017. Diabetes is also a major health problem, with a prevalence of 463 million globally and 29 million in the US. Obesity is thought to be a risk factor for both cardiovascular diseases (e.g. coronary heart disease, atrial fibrillation, ischemic stroke) and type 2 diabetes. Furthermore, obesity has continued to rise in the US population. Diet is an important risk factor for obesity and cardiometabolic outcomes, as well as a modifiable factor for intervention that could reduce disease risk. The contribution of specific dietary factors to cardiometabolic disease is poorly understood due to ethical, economical, and practical issues with randomizing a sufficient number of individuals to dietary interventions. Thus, we rely on observational data, which is prone to confounding by factors such as socioeconomic status and lifestyle choices. Because genes are fixed at conception and randomly assorted within a population, Mendelian Randomization (MR) mimics a randomized controlled trial on the basis of genetic makeup, allowing inference of causality. MR requires large-scale data, and with the availability of data such as that in the UK Biobank (> 500k participants), we can now use MR to address questions regarding dietary exposures that have so far been beyond our reach. I will combine observational, genetic, and MR analyses in the UK Biobank with external genome-wide association study (GWAS) meta-analyses of risk factors and cardiometabolic disease outcomes to answer important dietary questions that have so far evaded us.
I aim to elucidate the relationship of several dietary factors with cardiometabolic disease using observational and MR studies.
In Aim 1, I will characterize alcohol using observational association, standard MR, and non-linear MR analyses based on variation in the alcohol dehydrogenase gene.
In Aim 2, I will characterize coffee using observational analysis, perform GWAS to create a genetic risk score for coffee intake and subtypes of coffee, and use the resulting genetic risk scores for MR and non-linear MR analyses.
In Aim 3, I will characterize various dairy intake types using observational analysis and perform an MR analysis based on variation in the lactase persistence gene. In summary, the combination of observational, genetic, and MR analyses will allow us to characterize the risk associated with ubiquitous alcohol, caffeine, and dairy consumption in a much more meaningful way than correlations drawn from observational analyses alone. This has tremendous potential to influence the advice given by nutritionists and physicians to the hundreds of millions of people who suffer from or are at risk of cardiovascular disease, diabetes, and obesity.
Dietary factors contribute to cardiometabolic diseases which represent a huge worldwide disease burden. Using observational analysis and Mendelian randomization, the association and causality of alcohol, coffee, and milk in cardiometabolic disease will be quantified.
