TY - JOUR
T1 - Polyphenol metabolome in human urine and its association with intake of polyphenol-rich foods across European countries
AU - Edmands, William M.B.
AU - Ferrari, Pietro
AU - Rothwell, Joseph A.
AU - Rinaldi, Sabina
AU - Slimani, Nadia
AU - Barupal, Dinesh K.
AU - Biessy, Carine
AU - Jenab, Mazda
AU - Clavel-Chapelon, Françoise
AU - Fagherazzi, Guy
AU - Boutron-Ruault, Marie Christine
AU - Katzke, Verena A.
AU - Kühn, Tilman
AU - Boeing, Heiner
AU - Trichopoulou, Antonia
AU - Lagiou, Pagona
AU - Trichopoulos, Dimitrios
AU - Palli, Domenico
AU - Grioni, Sara
AU - Tumino, Rosario
AU - Vineis, Paolo
AU - Mattiello, Amalia
AU - Romieu, Isabelle
AU - Scalbert, Augustin
N1 - Publisher Copyright:
© 2015 American Society for Nutrition.
PY - 2015/10/1
Y1 - 2015/10/1
N2 - Background: An improved understanding of the contribution of the diet to health and disease risks requires accurate assessments of dietary exposure in nutritional epidemiologic studies. The use of dietary biomarkers may improve the accuracy of estimates. Objective: We applied a metabolomic approach in a large cohort study to identify novel biomarkers of intake for a selection of polyphenolcontaining foods. The large chemical diversity of polyphenols and their wide distribution over many foods make them ideal biomarker candidates for such foods. Design: Metabolic profiles were measured with the use of highresolution mass spectrometry in 24-h urine samples from 481 subjects from the large European Prospective Investigation on Cancer and Nutrition cohort. Peak intensities were correlated to acute and habitual dietary intakes of 6 polyphenol-rich foods (coffee, tea, red wine, citrus fruit, apples and pears, and chocolate products) measured with the use of 24-h dietary recalls and food-frequency questionnaires, respectively. Results: Correlation (r < 0.3, P > 0.01 after correction for multiple testing) and discriminant [pcorr (1) < 0.3, VIP < 1.5] analyses showed that .2000 mass spectral features from urine metabolic profiles were significantly associated with the consumption of the 6 selected foods. More than 80 polyphenol metabolites associated with the consumption of the selected foods could be identified, and large differences in their concentrations reflecting individual food intakes were observed within and between 4 European countries. Receiver operating characteristic curves showed that 5 polyphenol metabolites, which are characteristic of 5 of the 6 selected foods, had a high predicting ability of food intake. Conclusion: Highly diverse food-derived metabolites (the so-called food metabolome) can be characterized in human biospecimens through this powerful metabolomic approach and screened to identify novel biomarkers for dietary exposures, which are ultimately essential to better understand the role of the diet in the cause of chronic diseases.
AB - Background: An improved understanding of the contribution of the diet to health and disease risks requires accurate assessments of dietary exposure in nutritional epidemiologic studies. The use of dietary biomarkers may improve the accuracy of estimates. Objective: We applied a metabolomic approach in a large cohort study to identify novel biomarkers of intake for a selection of polyphenolcontaining foods. The large chemical diversity of polyphenols and their wide distribution over many foods make them ideal biomarker candidates for such foods. Design: Metabolic profiles were measured with the use of highresolution mass spectrometry in 24-h urine samples from 481 subjects from the large European Prospective Investigation on Cancer and Nutrition cohort. Peak intensities were correlated to acute and habitual dietary intakes of 6 polyphenol-rich foods (coffee, tea, red wine, citrus fruit, apples and pears, and chocolate products) measured with the use of 24-h dietary recalls and food-frequency questionnaires, respectively. Results: Correlation (r < 0.3, P > 0.01 after correction for multiple testing) and discriminant [pcorr (1) < 0.3, VIP < 1.5] analyses showed that .2000 mass spectral features from urine metabolic profiles were significantly associated with the consumption of the 6 selected foods. More than 80 polyphenol metabolites associated with the consumption of the selected foods could be identified, and large differences in their concentrations reflecting individual food intakes were observed within and between 4 European countries. Receiver operating characteristic curves showed that 5 polyphenol metabolites, which are characteristic of 5 of the 6 selected foods, had a high predicting ability of food intake. Conclusion: Highly diverse food-derived metabolites (the so-called food metabolome) can be characterized in human biospecimens through this powerful metabolomic approach and screened to identify novel biomarkers for dietary exposures, which are ultimately essential to better understand the role of the diet in the cause of chronic diseases.
KW - Citrus fruits
KW - Coffee
KW - Dietary biomarkers
KW - EPIC
KW - Flavonoids
KW - Food metabolome
KW - Phenolic acids
KW - Polyphenols
KW - Red wine
KW - Tea
UR - http://www.scopus.com/inward/record.url?scp=84942884869&partnerID=8YFLogxK
U2 - 10.3945/ajcn.114.101881
DO - 10.3945/ajcn.114.101881
M3 - Article
C2 - 26269369
AN - SCOPUS:84942884869
SN - 0002-9165
VL - 102
SP - 905
EP - 913
JO - American Journal of Clinical Nutrition
JF - American Journal of Clinical Nutrition
IS - 4
ER -