The long-term objective of this Program Project is to identify novel hormonal, dietary and genetic determinants of cancer risk (specifically breast, colorectal, and ovarian cancers) in v\/omen, with the aim of finding means for prevention and improved survival. The availability of questionnaire data, blood and urine samples and paraffin-embedded tumor tissue, coupled with up to 36 years follow-up, affords the opportunity to further understand the time course, as well as mechanisms of cancer development. Cross-cutting themes include the investigation of vitamin D and melatonin, energy balance and inflammatory pathways, as well as the examination of fatal outcomes and the molecular characterization of tumors. In Project 1, we will examine vitamin D, melatonin, and dietary intake (including 1-carbon nutrients and change in alcohol intake) in relation to breast cancer risk; further, risk factors for clinically relevant tumor subtypes and fatal disease will be assessed. We will assess the contribution of modifiable factors to incidence in women at high genetic risk. Further the role of energy balance in survival will be assessed. In Project 2, we will assess vitamin D and melatonin, including associations with tumor characteristics. The roles of insulin, adiponectin and insulin-like growth factors as well as inflammation markers in colon cancer incidence and survival will be assessed, in Project 3, we will evaluate vitamin D, melatonin, dietary factors (flavonoids, lactose and acrylamide) and plasma markers of inflammation in relation to ovarian cancer risk. Further, we will assess tumor markers in relation to ovarian cancer risk factors, as well as risk factors for fatal disease. In Project 4, we will extend and apply analytic methods that will improve our modeling of cumulative exposures that influence risk at different times in life, extend polytomous logistic regression models in a complex tumor biomarker setting, and improve evaluation of breast and colon cancer risk prediction models. This application is based on the Nurses' Health Study cohort, comprising 121,700 women who were 30 to 55 years of age in 1976. Through funding of the Cores, we maintain follow-up of the cohort and confirmation of incident cancers and deaths, oversee acquisition and use of biomarkers and provide leadership and data analysis. The Nurses' Health Study as funded through this Program Project contributes to the NCI sponsored Cohort Consortium, belongs to 12 other consortia and has served as the research platform for over 30 outside collaborators over the last 5 years.