We propose a cost-effective expansion of a population-based cohort study, the Atherosclerosis Risk in Communities (ARIC) Study, to encompass etiologic cancer research. A total of 15,000 middle-aged adults, blacks and whites, sampled from four U.S. communities were first examined in 1987-1989. Many known or potential risk factors for common cancers were carefully measured in the baseline examination, and biologic specimens (sera, plasma, blood cells, and DNA) were stored. Three subsequent examinations have been funded: 1990-1992, 1993-1995, and 1996-1998. Participants have reported cancer occurrence at every examination and have annually reported all hospitalizations, including those for cancer. About 450 cancers are estimated to have occurred by 1992, an average follow-up of four years. The ARIC Study is likely to continue long-term, so expanding it to include cancer endpoints is cost-effective. There are five specific aims for this three-year project. First, we propose in Year 01 to augment information on established risk factors for common cancers (breast, lung, colon, and prostate) by means of a phone interview. Second, we propose to expand ARIC's rudimentary system for finding incident cancer cases by linking, in Years 01 and 03, cohort identifiers to cancer registries, available in three of the four source communities, and by collecting in Years 01 through 03 cancer-related hospitalizations of cohort members. Third, a trained nurse abstractor will abstract information on the estimated 700 incident cancers that will have occurred between January 1987 and December 1994 (an average of follow-up of over six years). Fourth, in Year 03 we will test specific hypotheses related to 1) the independent etiologic role of body fat distribution, body mass, endogenous insulin, and lipid levels in breast cancer, and 2) whether chronic obstructive lung disease is a risk marker for lung cancer independent of smoking intensity. The fifth, and broader, aim is to establish cancer as a reliable endpoint in the ARIC cohort for future studies of genetic and environmental risk factors and their interactions. The proposed project, which takes advantage of an already established large population-based cohort, requires relatively little support and yet will contribute substantially to the etiologic research of common cancers during this project and far beyond it.
