Multiethnic Cohort Study of Diet and Cancer

描述（由申请人提供）：本申请请求继续资助1993- 1996年在夏威夷和洛杉矶建立的多民族队列研究，以探索饮食和其他生活方式因素与癌症的关系。该队列由21.5万多名男性和女性组成，主要来自非洲裔美国人、日本人、拉丁裔、夏威夷原住民和高加索人，在现有的队列研究中，其种族多样性和少数民族人口的代表性是独一无二的。在入学时，每个参与者都完成了一份26页的邮件调查问卷，其中包括广泛的定量饮食史，以及其他人口统计、医疗和生活方式信息。在一项校准研究中，我们收集了2000多名参与者的24小时饮食记录，该研究旨在校正营养摄入估计的测量误差。在当前周期（1998-2003）中，超过80%的参与者完成了一份简短的随访问卷，以更新选定的饮食和非饮食基线信息；在下一个周期中，将重复基线饮食调查问卷。7年后的外迁率很低（< 5%），支持使用计算机与夏威夷和加利福尼亚基于人群的癌症登记处联系，以确定事件病例；预计到2007年，乳腺癌、前列腺癌、结直肠癌和肺癌的病例将超过2.5万例。基于平均4年的随访，一些初步的饮食发现包括：酒精与乳腺癌呈正相关，乳制品与前列腺癌呈正相关，饱和脂肪与肺癌呈正相关；豆类与乳腺癌呈负相关，类胡萝卜素和某些蔬菜与肺癌呈负相关，膳食纤维和叶酸与结直肠癌呈负相关。在下一个周期中，更多的病例将使我们能够进一步根据种族和疾病阶段检查这些饮食关系，检查其他癌症部位（胰腺，膀胱，卵巢，子宫内膜），并为基因-饮食相互作用的分析做出贡献。此外，后续问卷调查的新信息和食品成分数据库的增强将用于直接测试黄酮类化合物和杂环胺与结直肠癌、肺癌、膀胱癌等癌症的关系。我们还将检验与使用外源性激素和乳腺癌、卵巢癌和结直肠癌有关的假设；非甾体抗炎药的使用以及结直肠癌、肺癌、乳腺癌和卵巢癌的发病率。方法学的目标是利用750名队列成员的生化测量，开发新的饮食暴露变量，以改进病因假设的测试。多种族队列研究的结果不仅有助于确定癌症的饮食和其他风险因素，还有助于更好地了解癌症发病率的种族差异的基础。

DESCRIPTION (provided by applicant): This application requests funding for continued support of the Multiethnic Cohort Study which was established in Hawaii and Los Angeles in 1993-96 to explore the relationship of diet and other lifestyle factors to cancer. The cohort is comprised of more than 215,000 men and women primarily of African American, Japanese, Latino, Native Hawaiian and Caucasian origin, and is unique among existing cohort studies in its ethnic diversity and representation of minority populations. At entry, each participant completed a 26-page mail questionnaire that included an extensive quantitative diet history, as well as other demographic, medical, and lifestyle information. Multiple 24-hour diet recalls were collected on more than 2,000 of the participants in a calibration study designed to permit correction of nutrient intake estimates for measurement error. In the current cycle (1998-2003), a brief follow-up questionnaire to update selected dietary and non-dietary baseline information was completed by more than 80 percent of the participants; in the next cycle, the baseline dietary questionnaire will be repeated. Out-migration rates after 7 years were low (< 5 percent), supporting the use of computer linkage with the population-based cancer registries in Hawaii and California to identify incident cases; more than 25,000 cases of breast, prostate, colorectum and lung cancer are expected by the year 2007. Based on an average 4 years of follow-up, some preliminary dietary findings included: positive association of alcohol with breast cancer, dairy products with prostate cancer, and saturated fat with lung cancer; and inverse association of legumes with breast cancer, carotenoids and certain vegetables with lung cancer, and dietary fiber and folate with colorectal cancer. In the next cycle, the larger numbers of incident cases will enable us to further examine these dietary relationships by ethnicity and stage of disease, to examine other cancer sites (pancreas, bladder, ovary, endometrium), and to contribute to analyses of gene-diet interactions. In addition, new information from the follow-up questionnaire and enhancements to the food composition database will be used to directly test the relationship of flavonoids and heterocyclic amines to colorectal, lung, bladder and other cancers. We will also test hypotheses related to use of exogenous hormones and cancers of the breast, ovary and colorectum; and use of NSAIDs and cancers oft he colorectum, lung, breast and ovary. A methodological aim, utilizing biochemical measurements on 750 members of the cohort, will develop new dietary exposure variables to improve testing of etiologic hypotheses. Findings from the Multiethnic Cohort Study should help not only to identify dietary and other risk factors for cancer, but also to better understand the basis for ethnic variations in cancer incidence.