Lipid Genotypes, Phenotypes, and Colorectal Polyps

Abstract

Studies linking cholesterol levels to colorectal neoplasia have been inconsistent. This dissertation aimed to clarify whether dyslipidemia is a risk factor for adenomatous or non-adenomatous colorectal polyps using newly-identified information on cholesterol genetics. We utilized epidemiologic data and biospecimens from a colonoscopy study conducted from 1998 to 2007 among enrollees of Group Health, a large healthcare system in Washington State. Participants were 518 non-advanced adenoma cases, 139 advanced adenoma cases, 380 non-adenomatous polyp cases, and 754 polyp-free controls. New data collected for this research included: 1) genotypes of 96 single-nucleotide polymorphisms identified from genome-wide association studies of low- and high-density lipoprotein cholesterol (LDL, HDL), triglycerides, and total cholesterol; 2) clinical cholesterol measurements from Group Health's laboratory records; and 3) information on lipid-controlling prescription drug use from Group Health's pharmacy records. Chapter 1 provides introductory results. Compared to those who reported no physician-diagnosis of hypercholesterolemia, those who reported untreated hypercholesterolemia had increased prevalence of advanced adenomas. This association was not observed for those who reported treated hypercholesterolemia. Statin use was more likely to be associated with a decreased prevalence of advanced adenomas at the highest observed pre-colonoscopy LDL levels. Chapter 2 describes a systematic review and meta-analysis of 18 studies that assessed cholesterol from blood at the time of endoscopy, including 6,645 adenoma cases and 21,335 polyp-free controls. Individuals with colorectal adenoma were more likely than controls to have higher total cholesterol, higher triglycerides, and lower HDL. Chapter 3 describes a validation study to assess the accuracy of self-reported hypercholesterolemia, which we found to be accurate with higher specificity than sensitivity. Chapter 4 compares phenotype-polyp associations to genotype-polyp associations estimated from Mendelian randomization analyses. Given that colorectal neoplasia shares many risk factors with cardiovascular disease, including obesity, physical inactivity, and smoking, Mendelian randomization is an attractive approach to help avoid problems with confounding and reverse causation. Dyslipidemia may be a marker of the type of adiposity and dietary exposures that promote neoplasia, but there was insufficient evidence to conclude that genetic susceptibility to dyslipidemia is associated with colorectal polyps. Chapter 5 summarizes primary study limitations and recommendations for future studies.