We propose to study prospectively the dietary etiologies of colorectal adenomas, precursors of the vast majority of colorectal cancers, in men and women from two large established cohorts, the Nurses' Health Study and the Health Professionals Follow-up Study. Participants in these studies (90,000 women and 52,000 men) have already completed a mailed self-administered semiquantititative food frequency questionnaire in 1986. The questionnaire has been intensively validated in subgroups of both cohorts. The cohort members report newly diagnosed diseases (including colon polyps) on mailed questionnaires every two years. The proposed study population will be restricted to subjects from both cohorts who have colonoscopy or sigmoidoscopy between 1986 and 1990, and will exclude persons who reported polyps before 1986. Cases are subjects with a first diagnosis of colorectal adenoma between 1986 and 1990, and non-cases are defined as subjects who had a endoscopic procedure negative for adenomas in the same time interval. We propose to obtain pathology reports to confirm the diagnosis of adenoma for all cases. We will also validate the pathologic diagnoses by reviewing a sample of the microscopic slides. Using nutrient intakes estimated from the 1986 dietary questionnaires, we plan to test the following hypotheses: high fiber intake reduces the risk and high fat intake increases the risk of developing colorectal adenomas. Other nutrients of interest that we will study are calcium, vitamins C and E, preformed vitamin, A, carotenoids and alcohol intake. We also have information on cigarette smoking, weight and height, exercise, and family history of colorectal cancer. Stratified and multivariate techniques will be used to quantify the age-adjusted relative risk of colorectal adenoma according to level of nutrient intake after controlling for other potentially important confounders. We project that 900 incident cases will be confirmed; this will provide 99% power to detect a relative risk of 1.5 between the high and low quintiles of specific nutrient intakes.
