Combined Impact of Lifestyle, Dietary and Metabolic Factors on Colorectal Cancer Incidence and Survival in European Populations: Implications for Cancer Risk Assessment and Risk Prediction

结直肠癌（CRC）是全球男性第三大常见癌症，女性第二大常见癌症，近55%的病例发生在发达国家，如西欧。迫切需要在欧洲男性和女性中开发CRC风险预测模型，以加强其预防。疾病频率与西方化水平的平行关系表明，生活方式和代谢因素可能参与了CRC的病因学。越来越多的证据，包括欧洲癌症和营养前瞻性调查（EPIC）的数据表明，行为因素，如酒精，吸烟，体育活动，身体肥胖，以及某些饮食和代谢因素与CRC风险较高有关。虽然这些因素的单独作用已被广泛研究，但对其联合作用知之甚少。然而，在真实的生活中，人们的孤立行为并不常见，因此多维数据分析方法将为探索疾病病因提供更多信息，并将流行病学研究结果转化为有意义的公共卫生信息。通过这种应用，我们建议探索生活方式，饮食和代谢因素对CRC发病率和生存率的影响，并开发实用的CRC风险评估和风险预测的公共卫生工具。更具体地说，我们将使用CRC风险因素的先验知识来组成生活方式和代谢生物标志物指标，并探索这些指标与CRC发病率和生存率的相关性。作为进一步的步骤，我们将开发一个CRC风险预测模型，并评估该模型在校准，区分和准确性方面的性能。最后，我们将探讨新建立的生物标志物和遗传数据是否可能改善已知CRC风险因素以外的风险预测。拟定的分析将纳入EPIC队列，该队列包括519，928名年龄在25-70岁之间的男性和女性，他们提供了饮食和生活方式信息、人体测量和血液样本（1992-2000年），并随访至2010年9月31日。在平均12年的随访中，确定了4，452例CRC事件（2，817例结肠和1，635例直肠）和1，107例CRC特异性死亡。该项目将利用EPIC队列的独特研究环境，并将受益于其强大的设计，大量的病例和各种风险因素的高质量流行病学数据的可用性。本项目中开发的风险预测模型最终可用于确定CRC的高危人群和制定干预策略。因此，该项目有望在临床实践和公共卫生中对CRC的一级预防产生直接影响。

Colorectal cancer (CRC) is the third most common cancer in men and the second in women worldwide with almost 55% of the cases occurring in developed countries, such as in Western Europe. Development of risk prediction models for CRC in European men and women is urgently needed to enhance its prevention. The parallel of disease frequency with the level of westernization points to the likely involvement of lifestyle and metabolic factors in the etiology of CRC. Mounting evidence, including data from the European Prospective Investigation into Cancer and Nutrition (EPIC) suggests that behavioral factors such as alcohol, smoking, physical activity, body fatness, and certain dietary and metabolic factors are individually related to a higher CRC risk. While individual roles of these factors have been extensively investigated, little is known about their joint effects. However, in real life it is uncommon that people practice isolated behaviors; therefore a multidimensional data analysis approaches would be more informative for exploring disease etiology, as well as for translating epidemiological findings into meaningful public health messages. With this application, we propose to explore the impact of combinations of lifestyle, dietary and metabolic factors on CRC incidence and survival, and to develop practical public health tools for CRC risk assessment and risk prediction. More concretely, we will use a-priori knowledge on CRC risk factors to compose lifestyle and metabolic biomarker indexes and explore associations of these indexes with CRC incidence and survival. As a further step, we will develop a CRC risk prediction model and evaluate the performance of this model in terms of its calibration, discrimination and accuracy. Finally, we will explore whether newly established biomarker and genetic data potentially improves risk prediction beyond known CRC risk factors. Proposed analyses will be embedded within the EPIC cohort comprising of 519,928 men and women, aged 25-70 years, who provided dietary and lifestyle information, anthropometric measurements and blood samples (1992-2000), and were followed up through September 31, 2010. Over a mean follow-up of 12 years, 4,452 incident CRC cases (2,817 colon, and 1,635 rectal) and 1,107 CRC-specific deaths were identified. The project will utilize the unique study setting of the EPIC cohort, and will profit from its strong design, large number of cases and availability of high-quality epidemiological data on variety of risk factors. The risk prediction model developed within this project could be ultimately used for identifying high risk groups and developing intervention strategies for CRC. Therefore, this project brings promises to be of direct impact for the primary prevention of CRC in clinical practice and public health.