Biomarkers of Dietary Variables Commonly Misreported

目前测量自由生活人群饮食摄入量的方法通常会导致能量摄入量（EI）平均低估10- 50%。数据提示，但不是结论性的，EI低报是特定于高能量密度的食品和饮料（即，高糖和/或脂肪）。我们和其他人的工作表明，饮食误报，即低报或多报，可能会导致饮食变量与体重指数（BMI）和慢性疾病风险因素的虚假统计关联，这反过来又可能导致关于过度饮食，肥胖和其他健康结果的潜在饮食贡献者的错误结论。生物标志物作为饮食变量的客观，独立的措施，被认为是常见的和选择性的误报将有助于验证哪些具体的饮食因素被误报的个人。在对PAR-07-259，改善饮食和身体活动评估的回应中，我们的总体假设是，饮食误报是特定于某些类型的食物和/或营养素;特别是，高糖和/或高脂肪的食物被低估，水果和蔬菜被高估。我们将招募n=181名不同种族的男性和女性，年龄在18-65岁之间，BMI为18.5-45 kg/m2，主要来自健康普渡大学人群-一个基于大学的健康生活方式计划，参与者超过9000人-但也来自其他来源。我们将测试我们的假设，通过评估一个新的组合（尿蔗糖+果糖的食物高糖）和建立（水果和蔬菜的血浆类胡萝卜素和血清叶绿醌）膳食生物标志物，报告的膳食摄入量，以及它们之间的关系以及与总研究样本（n=181）中健康结局的关系，与生理学上合理的EI报告者的子样本相比（目标n≥120;其中，根据先前建立的方法，如果个体的EI在通过双标记水测量的总能量消耗的±1 SD内，则将其视为合理）。我们希望这项研究的结果能提高我们对膳食评估方法中报告错误性质的理解，并最终有助于澄清膳食摄入量和健康结果的不同组成部分之间的关系。

Current methods for measuring dietary intake in free-living people typically result in an average under-estimation of energy intake (EI) of 10-50%. Data are suggestive, but not conclusive, that EI underreporting is specific to foods and beverages high in energy density (i.e., high in sugar and/or fat). Our work and that of others has shown that dietary misreporting, that is, underreporting or overreporting, may result in spurious statistical associations of dietary variables with body mass index (BMI) and chronic disease risk factors, which in turn may lead to incorrect conclusions about potential dietary contributors to overeating, obesity, and other health outcomes. Biomarkers that serve as objective, independent measures of dietary variables thought to be commonly and selectively misreported would help to verify which specific dietary factors are misreported by which individuals. In this response to PAR-07-259, Improving Diet and Physical Activity Assessment, our overall hypothesis is that dietary misreporting is specific to certain types of foods and/or nutrients; in particular, foods high in sugar and/or fat are underreported, and fruits and vegetables are overreported. We will recruit n=181 ethnically diverse men and women aged 18-65 y and BMI 18.5-45 kg/m2 primarily from the Healthy Purdue population--a university-based healthy lifestyle program with >9000 participants--but also from other sources. We will test our hypotheses by assessing a combination of novel (urinary sucrose + fructose for foods high in sugar) and established (plasma carotenoids and serum phylloquinone for fruits and vegetables) dietary biomarkers, reported dietary intakes, and their relationships with each other as well as with health outcomes in the total study sample (n=181) compared to the subsample of physiologically plausible EI reporters (target n≥120; where, according to previously-established methodology, an individual’s EI will be regarded as plausible if it is within ±1 SD of total energy expenditure measured by doubly labeled water). We expect the results of this study improve our understanding of the nature of reporting error in dietary assessment methodology, and ultimately help clarify relationships between different components of dietary intake and health outcomes.