Epidemiology provides a means of investigating the architecture of complex diseases by combining advances in technology with our current understanding of the epidemiology of disease. In this dissertation, epidemiologic methods are applied to further understand the etiology of colorectal cancer, childhood cancers, and congenital heart disease (CHD).

The low-penetrance genes that contribute to risk of familial colorectal cancers (CRCs) are mostly unknown. Two leading candidate regions resulting from the Molecular Epidemiology of Colorectal Cancer (MECC) study genome-wide association study included rs10210149 on chromosome 2q11.2-q12 and rs16931815 on chromosome 12p11.23. After excluding potential pathogenic mutations by Sanger sequencing, allele-specific expression analyses were performed for GPR45, STK38L, and TGFBRAP1. GPR45 was associated with a 27% increase in expression for each additional copy of the C allele of rs10210149 (p-trend = 0.01).

Common genetic variation and risk of congenital heart disease has not previously been studied. We investigated variation in ISL1, a marker of cells that contribute to specific developmental fields of the embryonic human heart, and risk of CHD in a two-stage case-control study. Eight genic and flanking ISL1 SNPs were significantly associated with complex CHD. A replication study analyzed the three SNPs within ISL1 (rs3762977, IVS1+17C>T, rs1017) and confirmed that genetic variation in ISL1 is associated with risk of CHD. Our results demonstrate that two different ISL1 haplotypes contribute to risk of CHD in white (Summary Odds Ratio (OR) =1.27, 95% Confidence Interval (CI) 1.09 – 1.48, P = 0.0018) and black/African American populations (Summary OR=1.57, 95% CI 1.07 – 2.30, P = 0.0216).

Linking epidemiologic approaches to cancer epidemiology and cardiovascular disease is stimulated by the well known association between selected forms of congenital heart disease and cancer. We conducted a retrospective cohort study of CHD at the Children's Hospital of Philadelphia (CHOP), showing that children with CHD demonstrated a 3.7-fold increase in the rate of pediatric cancer compared to the US population (Standardized Incidence Ratio (SIR) = 3.72, 95% CI = 1.53 – 9.04, p = 0.0037). Rates were higher for children with both syndromic (SIR=12.49, 95% CI 1.28 – 121.74, p=0.03) and non-syndromic (SIR=2.41, 95% CI 0.88 – 6.60, p=0.086) heart disease.