PROSPECTIVE STUDY OF DIET AND CANCER IN MEN (HEALTH PROFESSIONALS FOLLOWUP)

我们建议继续对51,529名男性卫生专业人员进行随访， 1986年，年龄在40岁至75岁之间，以解决一系列饮食假设 与大肠癌、前列腺癌、肺癌和黑色素瘤有关。我们 已经记录了在摄入量方面的巨大差异 这一人群中的营养素以及我们饮食的能力 1986年、1990年和1994年进行的问卷调查，以了解这一点 变种。对两年一次的后续调查问卷的答复至少是 94%。1987年，我们收集了腰围和臀围的数据，这将 将于1996年更新。1994年，我们收集并储存在液氮中， 来自18,000名参与者的血浆和DNA样本。进行的分析 在本供资期间，对以下几项作出了重要贡献 关于结肠癌和前列腺癌的长期假设，以及 还提出了几个新的、有希望的方向。 我们提出的具体目标建立、扩展和完善了我们的观察结果 已经做出了，部分使用生物样本来解决 一系列潜在的基因-饮食相互作用。在拟议的后续行动期间 我们预计总共有607例结肠癌和2164例结肠癌 前列腺癌。有了这些数据，我们将能够检查更长的时间 诱导期(这对结肠可能特别重要 癌症)，计算相对精确的效应估计，并检查 潜在的营养-营养相互作用。饮食记录的可得性 队列次样本中的数据和我们小组开发的统计方法 允许计算相对风险估计和可信区间 已修正测量误差。反复测量饮食，而不是 在任何其他大的男性队列中可用，将允许评估 长期膳食摄入量和饮食变化与糖尿病发病风险的关系 男性的重要癌症。

We propose to continue the follow-up of 51,529 male health professionals, aged 40 to 75 years in 1986, to address a series of dietary hypotheses related to cancers of the large bowel, prostate, lung, and melanoma. We have documented substantial variation in intake of a wide variety of nutrients in this population as well as the ability of our dietary questionnaire, administered in 1986, 1990, and 1994, to capture this variation. Response to biennial follow-up questionnaires has been at least 94%. In 1987 we collected data on waist and hip circumferences, which will be updated in 1996. In 1994 we collected, and stored in liquid nitrogen, plasma and DNA specimens from 18,000 participants. Analyses conducted during the current funding period have contributed importantly to several long-standing hypotheses regarding cancers of the colon and prostate, and have also suggested several new and promising directions. Our proposed specific aims build upon, extend, and refine observations we have already made, in part using the biological specimens to address a series of potential gene-diet interactions. During the proposed follow-up we anticipate a total of 607 cases of large bowel cancer and 2164 cases of prostate cancer. With this data we will be able to examine longer induction periods (which are likely to be particularly important for colon cancer), compute relatively precise estimates of effects, and examine potential nutrient-nutrient interactions. The availability of diet record data in a cohort subsample and statistical methods developed by our group allow the computation of relative risk estimates and confidence intervals corrected for measurement error. The repeated measurements of diet, not available in any other large cohort of men, will allow an assessment of the relation of longterm dietary intakes and change in diet to risks of important cancers in men.