Association of Nutritional and Biochemical Predictors of Body Fat Amount and Distribution with Cancer Risk

项目总结(见说明)： 与全身、外周或皮下脂肪相比，身体肥胖，特别是腹部、腹内和肝脏区域的脂肪，被认为具有更高的代谢风险。然而，依赖于体重和腰围替代指标的癌症流行病学研究，很可能低估了高危隔间脂肪块的影响，这种影响在少数族裔中发现的比例更高。我们假设，可以通过人口统计学、简单的人体测量和生活方式暴露变量(与暴露相关的血液生化和行为特征)的组合来预测体脂分布。事实上，我们在使用多种族队列(MEC)研究的可行性研究中发现，预测 腹部MRI测量的腹部脂肪间隔，除了BMI外，还可以考虑暴露变量时得到改善。在这项拟议的研究中，我们将首先评估饮食摄入、体力活动和认知进食的能量相关行为，并将分析2000名MEC受试者(10个性别民族各200人)的脂肪细胞因子、胰岛素/IGF、性类固醇激素、血脂和脂溶微量营养素的水平，这些受试者将再次接受DXA和腹部MRI测量的详细肥胖症表型的横断面研究。使用随机森林和其他类似的回归模型，我们将确定最佳的营养、生化和行为标记物 改善体重指数对以下肥胖表型的预测：DXA测量的总脂肪量和躯干/外周脂肪比，以及MRI测量的腹深和浅层皮下脂肪、内脏脂肪和肝脏脂肪。最后，我们将使用前瞻性收集的问卷数据和MEC中的生物标本，在嵌套病例对照研究中检验预测肥胖表型和顶级预测因素与乳腺癌(n=1217例)和结直肠癌(n=1379例)风险的相关性。这些数据将是研究暴露组、基因组、代谢组和肠道微生物组与身体脂肪分布和癌症风险的关系的POTS综合方法所必需的，因为它们 可能有助于确定行为干预的新途径。

PROJECT SUMMARY (See Instructions): Body fatness, in particular, in the abdominal, intra-abdominal and hepatic regions is known to carry a higher metabolic risk, compared to overall, peripheral or subcutaneous fat. Yet, the effect of fat mass in high-risk compartments, which is found in greater proportions in ethnic minorities, is likely to have been underestimated by cancer epidemiology research that relied on proxy measures of body mass and waist size. We hypothesize that body fat topography can be predicted by a combination of demographic, simple anthropometry, and lifestyle exposome variables (exposure-related blood biochemistry and behavior profiles). Indeed, we found in a feasibility study using the Multiethnic Cohort (MEC) study that the prediction of the abdominal fat compartments measured by abdominal MRI can be improved when considering exposome variables in addition to BMI. In the proposed study, we will first assess energetics-related behaviors of dietary intake, physical activity and cognitive eating and will analyze fasting blood levels of adipocytokines, insulin/IGF, sex steroid hormones, and lipids and lipid-soluble micronutrients among 2,000 MEC subjects (200 in each of 10 sex-ethnic groups) who will be recontacted for a cross-sectional study of detailed adiposity phenotypes measured by DXA and abdominal MRI. Using Random Forest and other similar regression models, we will identify the best nutritional, biochemical and behavioral markers that improve the prediction from BMI of the following adiposity phenotypes: total fat mass and trunk-to-periphery fat ratio measured by DXA and deep and superficial abdominal subcutaneous fat, visceral fat and liver fat measured by MRI. Finally, we will test the associations of the predicted adiposity phenotypes, as well as top predictors, for association with the risk of breast (n=1,217 cases) and colorectal cancers (n=1,379 cases) in nested case-control studies using the prospectively collected questionnaire data and biospecimens in the MEC. These data will be essential to this POTs integrated approach to study the relationships of the exposome, genome, metabolome and gut microbiome with body fat distribution and cancer risk since they may help to identify new avenues for behavioral interventions.