Data Sheet 1_Urinary metabolites as biomarkers of dietary intake: a systematic review.pdf

BackgroundCurrent diet assessment tools, such as food frequency questionnaires, may result in misclassification bias from measurement error and misreporting. These limitations can be mitigated by diet-related biomarkers in urine specimens, an emerging approach to characterize dietary intake.
ObjectiveWe conducted a systematic review to identify urinary biomarkers with utility in accurately assessing dietary intake, including individual foods and food groups.
MethodWe retrieved studies from 2000 to 2022 from databases including Embase, CINAHL, Cochrane, and PubMed. Data extraction from included articles was conducted by two independent reviewers for cross validation. Articles identifying urinary biomarkers in relation to food groups/items with adult populations were included and were evaluated for bias using the Joanna Briggs Institute Critical Appraisal.
ResultsA total of 65 articles were included and categorized as biomarkers of fruit (n = 13), vegetables (n = 5), aromatics (n = 5), fruits and vegetables (n = 3), grains/fiber (n = 5), dairy (n = 3), soy (n = 10), coffee/cocoa/tea (n = 9), alcohol (n = 6), meat and proteins (n = 6), nuts/seeds (n = 3), and sugar and sweeteners (n = 4). Results expanded the context to which metabolites of foods were compared across similar and dissimilar food groupings. Plant-based foods were often represented by polyphenols, while others were distinguishable by innate food composition, such as sulfurous compounds in cruciferous vegetables or galactose derivatives in dairy.
ConclusionCurrent evidence suggests urinary biomarkers may have utility in describing intake of broad food groups, such as citrus fruits, cruciferous vegetables, whole grains, and soy foods, but may lack the ability to clearly distinguish individual foods. These findings indicate the potential of urinary biomarkers to monitor changes in dietary patterns. The improvement of diet assessment methodology is a key step toward strengthening research data validity and accurately measuring outcomes in chronic disease management.
Systematic review registrationhttps://www.crd.york.ac.uk/PROSPERO/view/CRD42022308255, Prospero CRD42022308255.

研究背景：当前的膳食评估工具，例如食物频率问卷（food frequency questionnaires），可能因测量误差与报告偏差产生分类错误偏倚。而尿液样本中的膳食相关生物标志物可缓解这些局限，这是一种用于表征膳食摄入情况的新兴研究方法。
研究目的：本研究开展一项系统综述，旨在筛选可用于准确评估膳食摄入（包括单一食物及食物类别）的尿液生物标志物（urinary biomarkers）。
研究方法：本研究检索了2000年至2022年间Embase、CINAHL、Cochrane图书馆及PubMed数据库中的相关研究。由两名独立评审员对纳入文献进行数据提取以开展交叉验证。纳入针对成人人群、探讨与食物类别/单一食物相关的尿液生物标志物的文献，并采用乔安娜·布里格斯研究所批判性评价量表（Joanna Briggs Institute Critical Appraisal）对其偏倚风险进行评估。
研究结果：最终纳入65篇文献，按食物类别分为：水果类生物标志物（n=13）、蔬菜类（n=5）、芳香类食物（n=5）、果蔬类（n=3）、谷物/膳食纤维类（n=5）、乳品类（n=3）、大豆类（n=10）、咖啡/可可/茶类（n=9）、酒类（n=6）、肉类与蛋白质类（n=6）、坚果/籽类（n=3）以及糖类与甜味剂类（n=4）。本研究结果拓展了不同相似及非相似食物类别间食物代谢物对比的研究范畴。植物性食物的生物标志物多为多酚类物质，而其他类别则可通过食物固有成分进行区分，例如十字花科蔬菜中的含硫化合物，或乳品类中的半乳糖衍生物。
研究结论：现有证据表明，尿液生物标志物可用于表征大类食物的摄入情况，例如柑橘类水果、十字花科蔬菜、全谷物及大豆食品，但难以清晰区分单一食物。本研究结果提示尿液生物标志物具备监测膳食模式变化的潜力。优化膳食评估方法学，是提升研究数据效度、精准衡量慢性病管理结局的关键环节。
系统综述注册信息：https://www.crd.york.ac.uk/PROSPERO/view/CRD42022308255，注册编号：Prospero CRD42022308255。