ORIGINAL ARTICLE
Urinary excretion of 13 dietary flavonoids and
phenolic acids in free-living healthy subjects –
variability and possible use as biomarkers of
polyphenol intake
LI Mennen
1, D Sapinho
1, H Ito
2, P Galan
1, S Hercberg
1and A Scalbert
21U557 Inserm, U1125 Inra, Cnam, Paris 13/CRNH I-d-F, 74 rue Marcel Cachin, Bobigny, France and2Unite´ de Nutrition Humaine,
INRA, Centre de Recherche de Clermont-Theix, Saint-Gene`s-Champanelle, France
Objective: Estimation of dietary intake of polyphenols is difficult, due to limited availability of food composition data and bias inherent to dietary assessment methods. The aim of the present study was to evaluate whether we could detect polyphenols and their metabolites in a spot urine sample in a free-living human population and whether it was related to those observed in 24-h urine samples, for potential use as a biomarkers of polyphenol intake.
Subjects: Four 24-h urine samples and two spot urine samples were collected from 154 participants of the SU.VI.MAX cohort (a randomized primary-prevention trial evaluating the effect of daily antioxidant supplementation on chronic diseases) in two separate studies over, respectively, a 7- and 2-day periods. Thirteen polyphenols and metabolites (chlorogenic acid (CGA), caffeic acid (CA), m-coumaric acid (mCOU), gallic acid (GA), 4-O-methylgallic acid (MeGA), quercetin (Q), isorhamnetin (MeQ), kaempferol (K), hesperetin (HESP), naringenin (NAR), phloretin (PHLOR), enterolactone (ENL) and enterodiol (END) were measured using HPLC-ESI-MS-MS.
Results: Correlations between the urinary excretion levels were observed. The most significant were explained by metabolic filiations (CGA/CA, CA/mCOU, GA/MeGA, Q/MeQ, NAR/PHLOR, ENL/END) or co-occurrence in a same food source (NAR/ HESP). Concentrations in spot samples correlated with those in 24-h urine sample (Po0.02, except for CA and for MeQ). Intra-individual variations were smaller than inter-Intra-individual variations for all polyphenols (Po0.01) except for MeGA and for PHLOR. Conclusion: These results show that these polyphenols and metabolites are useful biomarkers for polyphenol intake.
European Journal of Clinical Nutrition (2008) 62, 519–525; doi:10.1038/sj.ejcn.1602744; published online 4 April 2007
Keywords: polyphenols; flavonoids; phenolic acids; biomarkers of intake; urine
Introduction
Polyphenols are compounds with high-antioxidant proper-ties and are probably the most abundant antioxidants in
our diet (Scalbert and Williamson, 2000). Both experi-mental and epidemiological evidence support a role of polyphenols in the prevention chronic diseases and
more particularly cardiovascular diseases and cancers
(Scalbert et al., 2005). Polyphenol intake has been related to disease in epidemiological studies, and especially inverse associations with cardiovascular risk have been observed (Hertog et al., 1993; Knekt et al., 1996; Hirvonen et al., 2001).
Many different types of polyphenols can be found in plant foods, one plant often containing more than one type of polyphenols and progress in epidemiological research on the relation between polyphenol consumption and disease is largely hampered by the lack of complete food composition tables. Although some food composition data are available for polyphenols, especially on the USDA website, content
Received 28 April 2006; revised 6 February 2007; accepted 19 February 2007; published online 4 April 2007
Correspondence: Dr A Scalbert, Unite´ de Nutrition Humaine, INRA, Centre de Recherche de Clermont-Ferrand/Theix, 63122 Saint-Gene`s-Champanelle, France.
E-mail: scalbert@clermont.inra.fr Guarantor:A Scalbert.
Contributors: LIM designed the study, participated in the statistical analyses and wrote the paper, DS performed the data handling and the statistical analyses, HI developed the analytical methods and performed the polyphenol analyses, PG and SH designed and supervised the SU.VI.MAX study and AS participated in the study design and initiated the development of the analytical methods. All contributors participated in the writing of the paper.
values for some polyphenol types or foods only consumed in specific countries are still missing. Furthermore, there is no gold standard for collection of dietary intake data, which is always subject to a certain under- or overestimation (Block, 1982; Bingham, 1991). The use of valid biomarkers for intake may be of help where the estimation of dietary intake is particularly difficult (Bingham, 2002); the advantage of biomarkers in dietary assessment being that their random errors are truly random and not dependent on those involved in dietary questionnaires (Kaaks, 1997). Measure-ments of compounds in spot urine samples are ideal potential biomarkers as they are relatively easy to obtain in large epidemiological studies.
The aim of the present study was, therefore, to evaluate whether we could detect polyphenols and their metabolites in a spot urine sample in a free-living human population and whether it was related to those observed in 24-h urine samples, for potential use as a biomarker of polyphenol intake. An HPLC-ESI-MS-MS was developed to analyse 13 different phenolic compounds in urine samples (Ito et al., 2005). It has been first validated using urine samples from controlled clinical trials and is applied here to urine samples collected from free-living subjects.
Subjects and methods
Subjects
Subjects were participants of the SU.VI.MAX study, a
randomized double-blind placebo-controlled
primary-prevention trial evaluating the effect of daily antioxidant supplementation (vitamin C, vitamin E, b-carotene, sele-nium and zinc) at nutritional doses on the incidence of cancer and ischemic heart disease. The cohort comprised women in the age of 35–60 years (mean: 46.4, s.d.: 6.7) and men in the age of 45–60 years (mean: 51.1, s.d.: 4.7) at baseline in 1994, and none of them used vitamin supple-ments other than those under study. In total, 13 077 subjects were included and were followed up for 8 years. Details on recruitment and study design are described elsewhere (Hercberg et al., 2004). SU.VI.MAX subjects living in the Parisian area were invited in 2002 to participate in a satellite protocol on the validation of a salt questionnaire, for which 154 subjects participated (study 1). The next year out of these 154 subjects, 53 were asked to and completed correctly another protocol to evaluate the measurement of polyphe-nols in urine as a biomarker for polyphenol intake (study 2) and were included in the present analyses. Of them, 31 were women and 22 were men and the mean age was 58 years at the time of the protocol.
The SU.VI.MAX study has been approved by the ethical committee for studies on human subjects (CCPPRB No. 706) of Paris-Cochin Hospital, and the ‘Comite´ National Informa-tique et Liberte´’ (CNIL No. 334641), which advocates that all medical information is confidential and anonymous.
Data collection
The first study included a 7-day protocol during which subjects collected 24-h urine on days 2, 4 and 6. The subjects indicated the total urine volume for each collection and kept a tube of each mixture, which were later collected by a dietician. No validation of the 24-h urine collection was performed for this study.
For the second study, subjects were first visited by a trained dietician to explain the 2-day protocol. The morning of the first day the subjects started a 2-day dietary record. The morning of the second day, the first-spot urine sample (spot 1) was collected. The collection of the 24-h urine started right after the collection of the spot sample and lasted until the following day at the same hour. During the 24-h period, three p-aminobenzoic acid tablets were taken, one after the collection of the spot sample, one at 1600 and one at 2300 to check the completeness of the 24-h urine collection. Only those subjects for whom a recovery of 85% or more was measured (53 subjects) were included for the subsequent analyses. The second-spot sample (spot 2) was taken in the morning of the third day and was thus part of the 24-h urine collection. On the first day, the subjects did not eat or drink after 2300 and they brought the three tubes with the collected urine (two spot and one tube taken from the mixed 24-h urine) to the study centre after the completion of the 24-h urine collection. The tubes were stored at 201 until measurement of the polyphenols.
Polyphenol measurements
An HPLC-ESI-MS-MS offering a high selectivity of detection for a wide range of phenolic compounds and short run times was developed (Ito et al., 2005). In brief, urine samples (250 ml) were supplemented with an internal standard (taxi-folin), incubated with b-glucuronidase and sulfatase (from Helix pomatia, Sigma G0876; Sigma, St Louis, MO, USA), and extracted with ethyl acetate. The organic extracts were redissolved in 25% aqueous methanol and injected into the HPLC-ESI-MS-MS system using a short Zorbax Eclipse XDB-C18 (2.1 i.d. 30 mm, 3.5 mm, Agilent, Massy, France) and an API-2000 (Applied Biosystems, Canada) mass spectro-meter. A 4-min gradient of water–formic acid (100:0.1) and acetonitrile–water–formic acid (95:5:0.1) was applied. The whole cycle including elution and reequilibration of the column did not exceed 6 min per sample. Mass detection was carried out with negative ionization in multiple reaction monitoring mode.
Calibration curves were prepared by spiking blank urine with aliquots of standard mixture solutions with duplicated injections at each concentration level. Determination of catechin and epicatechin was not included in the present study due to an insufficient sensitivity of the mass spectro-meter for these compounds and a too low concentration in the urine samples.
Statistical analysis
Mean, median, 10th and 90th percentile for all polyphenols and metabolites were calculated for measurements of the second study. Spearman’s rank correlation coefficients were calculated to evaluate the relation between the spot and 24-h urine sample and between different polyphenols. Kappa coefficients were calculated to evaluate the agreement across tertiles of polyphenol concentrations for the different samples. This analysis provides information on the level of misclassification when data are treated in tertiles, as is often caried out in statistical analyses of large epidemiological studies. Intra- and inter-individual variations were calculated with analyses of variance, using data of the 24-h urine collection of both the first and the second study (four collections in total per subject).
Results
Thirteen phenolic compounds were analysed by HPLC-ESI-MS-MS in urine samples collected from 53 free-living subjects. Three types of compounds were found in urine: compounds found in their native form after deconjugation with enzymes (chlorogenic acid, caffeic acid, gallic acid, quercetin, kaempferol, hesperetin, naringenin and phloretin), methylated metabolites formed in the host tissues (4-O-methylgallic acid and isorhamnetin) and microbial metabo-lites formed in the colon (m-coumaric acid, enterolactone and enterodiol).
In the 24-h urine samples, enterolactone was found in the highest quantity, followed by caffeic acid and naringenin, whereas lower amounts of isorhamnetin, quercetin and kaempferol were found (Table 1).
Analysis of the levels of urinary excretion of the different phenolic compounds revealed the following correlations: chlorogenic acid correlated well with caffeic acid, quercetin and isorhamnetin, whereas caffeic acid also correlated with m-coumaric acid, isorhamnetin, kaempferol and naringenin (Table 2). Gallic acid correlated with 4-O-methylgallic acid, isorhamnetin and naringenin. Phloretin correlated with
isorhamnetin, kaempferol and naringenin. Hesperetin
correlated with naringenin and 4-O-methylgallic acid. Enterodiol correlated with enterolactone, quercetin and isorhamnetin.
The concentration of polyphenols in the first- and the second-spot samples all correlated well with the concentra-tion in the 24-h urine sample, except for caffeic acid in both spot samples and for isorhamnetin in the first-spot samples (Table 3). The kappa coefficient was statistically
significant (Po0.05) for all correlations between the
concentrations in the 24-h urine sample and in the second-spot sample. This was also the case for most of the polyphenols measured in the first-spot samples, except for kaempferol and naringenin. As expected, correlations were stronger for the second spot urine samples, which were part of the 24-h urine samples. The intra-individual variation
was smaller than the inter-individual variation for all polyphenols except for 4-O-methylgallic acid and for phloretin (Table 4).
Table 1 Mean (s.d.), minimum and maximum levels of excretion of polyphenols in 24-h urine (mmol/day) and in two consecutive spot samples (mmol/l) in a group of free-living subjects
Polyphenol Urine sample n Mean Median 10th percentile 90th percentile Chlorogenic acid 24 h 53 4.2 0 0 12.7 Spot 1 53 1.7 0 0 3.4 Spot 2 53 6.6 4.5 1.3 15.3 Caffeic acid 24 h 53 11.4 9.8 4.2 17.9 Spot 1 53 2.4 2 0 4.7 Spot 2 53 6.2 5.5 3.7 9.6 m-Coumaric acid 24 h 53 5.4 1.9 0 12.9 Spot 1 53 2.2 1 0 6.9 Spot 2 53 2.7 1.1 0 5.4 Gallic acid 24 h 53 1.6 0 0 5.4 Spot 1 53 0.8 0 0 1.9 Spot 2 53 1.1 0.9 0 2.9 4-O-Methylgallic acid 24 h 53 6.1 0 0 21.4 Spot 1 53 3.4 0 0 12.6 Spot 2 53 4.4 0 0 10.3 Quercetin 24 h 53 0.7 0.6 0 1.4 Spot 1 53 0.4 0.2 0 1 Spot 2 53 0.4 0.3 0 1 Isorhamnetin 24 h 53 0.5 0.4 0 1 Spot 1 53 0.3 0.2 0 0.5 Spot 2 53 0.3 0.2 0 0.7 Kaempferol 24 h 53 0.8 0 0 4.1 Spot 1 53 0.4 0 0 1.7 Spot 2 53 0.4 0 0 2.1 Hesperetin 24 h 53 3 0 0 2.2 Spot 1 53 1 0 0 2.3 Spot 2 53 1.8 0.4 0 5.2 Naringenin 24 h 53 9.9 2.1 0.3 15 Spot 1 53 4.1 1 0.3 11.4 Spot 2 53 5.6 1.6 0.4 13.5 Phloretin 24 h 53 0.7 0.2 0 2.2 Spot 1 53 0.5 0.1 0 1.3 Spot 2 53 0.4 0.2 0 1 Enterolactone 24 h 53 23.9 15.6 5.4 39.1 Spot 1 53 14 10.1 3.9 22.5 Spot 2 53 10.7 8.3 3.2 23.1 Enterodiol 24 h 53 2 0.9 0.4 3.2 Spot 1 53 1.1 0.6 0.2 2.2 Spot 2 53 0.8 0.4 0.2 1.2
Data from study 2.
Discussion
The use of tandem mass spectrometry to estimate poly-phenols in urine implies that a range of polypoly-phenols is selected a priori. Thirteen phenolic compounds were selected for their widespread occurrence in the human diet, and for
being representative of the main types of polyphenol classes (Ito et al., 2005). It is therefore possible that other compounds other than those measured here are also suitable as biomarkers for polyphenol intake. Isoflavones were not included due to the low consumption of soy-containing products, the main dietary source of isoflavones, in France. Catechins were not included due to too limited sensitivity
Table 2 Spearman’s rank correlation coefficients between urinary excretion levels of 13 polyphenols in 24-h urine samples (r, P) collected from free-living subjects
Polyphenol CGA CA mCOU GA MeGA Q MeQ K HESP NAR PHLOR ENL END CGA 1 0.5 0.22 0.15 0.15 0.35 0.28 0.26 0.13 0.08 0.2 0.05 0.17 0.0001 0.12 0.3 0.3 0.01 0.04 0.07 0.34 0.56 0.15 0.72 0.21 CA 1 0.32 0.25 0.09 0.23 0.28 0.34 0.04 0.31 0.19 0.15 0.13 0.02 0.07 0.53 0.09 0.04 0.01 0.76 0.03 0.16 0.27 0.34 mCOU 1 0.009 0.002 0.11 0.05 0.14 0.03 0.02 0.17 0.14 0.06 0.95 0.99 0.45 0.74 0.3 0.83 0.89 0.23 0.32 0.65 GA 1 0.75 0.12 0.36 0.02 0.25 0.28 0.29 0.07 0.13 o0.0001 0.4 0.009 0.88 0.07 0.04 0.04 0.62 0.35 MeGA 1 0.11 0.31 0.02 0.28 0.19 0.29 0.08 0.16 0.45 0.03 0.89 0.04 0.18 0.04 0.56 0.24 Q 1 0.83 0.44 0.06 0.25 0.22 0.09 0.39 o0.0001 0.001 0.7 0.07 0.12 0.52 0.004 MeQ 1 0.47 0.22 0.49 0.45 0.07 0.4 0.0004 0.12 0.0002 0.0006 0.64 0.003 K 1 0.1 0.1 0.39 0.14 0.25 0.5 0.48 0.004 0.33 0.07 HESP 1 0.41 0.12 0.01 0.02 0.002 0.4 0.93 0.89 NAR 1 0.51 0.06 0.12 o0.0001 0.69 0.37 PHLOR 1 0.09 0.03 0.54 0.84 ENL 1 0.42 0.002
Abbreviations: CA, caffeic acid; CGA, chlorogenic acid; END, enterodiol; ENL, enterolactone; GA, gallic acid; HESP, hesperetin; K, kaempferol; mCOU, m-coumaric acid; MeGA, 4-O-methylgallic acid; MeQ, isorhamnetin; NAR, naringenin; PHLOR, phloretin; Q, quercetin.
Data from study 2, n ¼ 53. Bold figures indicate significative correlations (Po0.05).
Table 3 Spearman rank correlation coefficients (r,p) between poly-phenol concentrations in 24-h and spot urine samples in free-living subjects
Polyphenol Spot day 1 Spot day 2
r P r P
Chlorogenic acid 0.37a 0.006 0.42a 0.002 Caffeic acid 0.23 0.10 0.04 0.75 m-Coumaric acid 0.64a o0.0001 0.64a o0.0001 Gallic acid 0.29a 0.03 0.57a o0.0001
4-O-Methylgallic acid 0.28a 0.04 0.32a 0.02 Quercetin 0.40a 0.003 0.60a o0.0001 Isorhamnetin 0.05 0.74 0.45a 0.0008 Kaempferol 0.27 0.05 0.66a o0.0001 Hesperetin 0.5a 0.01 0.57a o0.0001 Naringenin 0.27 0.05 0.74a o0.0001 Phloretin 0.33a 0.004 0.65a o0.0001
Enterolactone 0.74a o0.0001 0.71a o0.0001 Enterodiol 0.62a o0.0001 0.69a o0.0001
Data from study 2, n ¼ 53.
a
P kappa coefficiento0.05.
Table 4 Inter- and intra-individual variations (mmol/day) in the excretion levels of 13 different polyphenols in 24 h urine samples collected from free-living subjects
Polyphenol Inter Intra P for difference Chlorogenic acid 8.5 4.8 o0.0001 Caffeic acid 5.9 4.2 o0.0001 m-Coumaric acid 18.1 15.7 0.01 Gallic acid 2.2 1.5 o0.0001 4-O-Methylgallic acid 4.2 4.6 0.92 Quercetin 1.1 0.7 o0.0001 Isorhamnetin 15.5 10.6 o0.0001 Kaempferol 2.8 2.4 0.006 Hesperetin 18.2 13.6 o0.0001 Naringenin 14.4 10.3 o0.0001 Phloretin 13.5 14.7 0.90 Enterolactone 29.5 16.2 o0.0001 Enterodiol 25.7 19.6 o0.0001
Data originate from 154 subjects and a total of 556 urine samples with three or four samples per subject. Data from studies 1 and 2.
with the method developed for analysis. Using the same method, they have been estimated in a clinical trial in samples collected after intake of polyphenol-rich beverages (Ito et al., 2005). However, attempts to detect catechins with the same method in the urine of free-living subjects were unsuccessful.
The relative excretion of polyphenols in urine depends on their intake and on their bioavailability (Manach et al., 2004, 2005) and concentration values in 24-h urine differ according to polyphenols. They can be compared to those published previously (Table 5). Coffee is one of the main dietary source of chlorogenic acid. Mean urinary excretion of chlorogenic acid observed in the cohort is close to that observed after consumption of about two cups of instant coffee (Ito et al., 2005). Gallic acid is present in food either in a free form or esterified to flavanols. Comparison of gallic acid and 4-O-methylgallic acid recovery in urine after ingestion of either wine or tea suggested a low urinary recovery when gallic acid is ingested in an esterified form (Ito et al., 2005). Urinary excretion of gallic acid and 4-O-methylgallic acid in the present cohort is about six times higher than that observed after consumption of one cup of tea (Ito et al., 2005). It is also higher than that observed after 4 weeks of regular tea consumption (Hodgson et al., 2000). This suggests that the subjects ingested another dietary source of gallic acid which could be wine, known to contain higher concentration of the more easily absorbed free gallic acid.
Flavonols (quercetin, isorhamnetin and kaempferol), fla-vanones (naringenin and hesperetin) and dihydrochalcones (phloretin) originate mainly from the consumption of fruits and vegetables. Except for hesperetin, excretion levels of the other flavonoids are somewhat higher than those observed previously in two Danish studies, in which the subjects ingested either their habitual diet or a diet rich in fruit and vegetables (Table 5). Our average concentration of
entero-lactone also seems high as compared to those reported in several other studies (reviewed in Horn-Ross et al. (1997)) and is more comparable to the one in women following a macrobiotic diet (Adlercreutz et al., 1986). It is known that a French population, in general, eats more fruits and vegeta-bles than a general population in Anglo-Saxon countries, were most of the previous studies have been carried out, and this may be one reason for the higher values (Beer-Borst et al., 2000).
Some correlations were found between the excretion levels of the different phenolic compounds. Strongest correlations are explained by metabolic filiations such as for chlorogenic acid and caffeic acid (hydrolysis) (Gonthier et al., 2003), gallic acid and 4-O-methylgallic acid, quercetin and iso-rhamnetin (methylation), enterolactone and enterodiol (hydrolysis and reduction) (Borriello et al., 1985). We also reported in a previous study the excretion of phloretin after consumption of citrus juice and suggested that phloretin is a metabolite of naringenin formed by reductive opening of the heterocycle (Ito et al., 2005). This explains the high correlation observed here between these two compounds. m-Coumaric acid is also weakly correlated to caffeic acid and is formed by microbial dehydroxylation of the latter in the colon (Gonthier et al., 2003).
Correlations between other polyphenols can be explained by co-occurrence in a given food. This is the case for the correlation between hesperetin and naringenin (both found in orange and orange juice; Ito et al. (2005). Citrus juice also contains quercetin (Belajova and Suhaj, 2004) and this may explain the correlations observed between naringenin and isorhamnetin. Quercetin or isorhamnetin are correlated to enterodiol, the former being found in various vegetables and the latter being produced from various vegetables (Kirkman et al., 1995). Kaempferol and quercetin originate in plants from a common precursor (dihydrokaempferol) and their occurrence in plants may be linked and explain the observed
Table 5 Comparison of mean values of 24-h urinary polyphenol concentrations (mmol/L) in different studies Reference Our study Nielsen et al.
(2002) Krogholm et al. (2004) Noroozi et al. (2000) de Vries et al. (1998) DuPont et al. (2002) Horn-Ross et al. (1997); Adlercreutz et al. (1986) Hodgson et al. (2000)
Diet Habitual Habitual HFLV LFLV HFLV Habitual Habitual After 1.1 L cider
Habitual/Macrobiotic After tea consumption. 4-O-Methylgallic acid 6.1 1.1 Quercetin 0.7 0.1 0.2 0.2 0.2 0.2 0.1 0.9 (HFLV) 0.9 (after tea) Isorhamnetin 0.5 0.04 0.06 0.07 0.1 Kaempferol 0.8 0.2 0.3 0.3 0.5 Hesperetin 3.0 2.5 3.7 1.1 2.6 Naringenin 9.9 2.6 2.6 1.6 2.4 Phloretin 0.7 0.3 0.5 0.3 0.4 3.8 Enterolactone 23.9 0.3-20 / 18
Abbreviations: LFLV, low flavonoid diet; HFLV, high flavonoid diet.
correlation between kaempferol on one side and quercetin and its O-methylated metabolite isorhamnetin on the other side. Similar co-occurrences in foods may explain other weak correlations observed between some flavonoids (quercetin, kaempferol and naringenin) and caffeic acid, chlorogenic acid or gallic acid.
The results of the present study show a good correlation between polyphenols measured in a spot sample and in a 24-h urine sample collected during a same day, indicating that a spot sample may be used to assay potential biomarkers of polyphenol intake. The only polyphenol for which we did not find a relation between the concentration measured in the spot sample and in that measured in the 24 urine sample is caffeic acid. We have no explanation for this lack of association, but it means that caffeic acid measured in a spot sample is probably not suitable for use as a biomarker. As far as we know no other study has published correlations between polyphenols concentrations measured in a 24-h urine sample and a spot sample. Krogholm et al. (2004) evaluated flavonoids in both a 24-h urine and a spot urine sample, but did not correlate the two. Weaker correlations were observed between polyphenols measured in a spot sample collected the day before the collection of the 24-h urine sample. This is expected due to the irregular consump-tion of various polyphenol food sources over successive days. Intra-individual variation can be due to variation in the measurement and/or due to variation in the dietary intake of the individuals. Intra-individual variations are higher for some polyphenols. Variations observed for compounds like chlorogenic acid and gallic acid are relatively low. This may be due to the regular consumption by French adults of the studied population of coffee and wine, respectively, main dietary sources of these two polyphenols. The higher intra-individual variations observed for other polyphenols could be due to the more irregular consumption of their main dietary source (citrus juice and fruits for hesperetin and naringenin, or apple for phloretin) or to their microbial origin (enterolactone and enterodiol).
The higher ratio of inter- over intra-individual variation for all measured compounds, except for 4-O-methylgallic acid and phloretin suggests that they are potentially useful biomarker of polyphenol intake in an epidemiological setting. Indeed when in nutritional epidemiological studies intra-individual variation of the nutritional variable is lower than the inter-individual variation, it becomes possible to detect a relation between the variable and the disease, or markers of disease of interest.
In conclusion, chlorogenic acid, m-coumaric acid, gallic acid, quercetin, isorhamnetin, kaempferol, hesperetin, nar-ingenin, enterolactone and enterodiol measured in spot urine samples are potentially useful biomarkers for poly-phenol intake. Few authors have examined so far correla-tions between polyphenol intake and polyphenol excretion in urine in free-living populations (Maskarinec et al., 1998; Lampe et al., 1999; Hodgson et al., 2004). We recently showed that urinary polyphenols correlate with intake of
various polyphenol-rich food and beverages (Mennen et al., 2006). Further research evaluating the relation of the concentration of these polyphenols in urine with dietary polyphenol intake is necessary to validate their use as biomarkers in large epidemiological studies.
Acknowledgements
This project received a grant from the French Ministry of Research (N1 02 P 0681). Special thanks to Unilever Bestfoods France (UBF), the ‘Centre d’Information Scientifique The´ & Sante´ de Lipton’, UBF and the Agence Pour la Recherche et l’Information en Fruits et Le´gumes frais (APRIFEL) for financial support. We thank the Dunn Human Nutrition Unit (UK) who provided us the PABA tablets for the study and Sandrine Bertrais for revising the article, and also Ministry of Agriculture, Unilever-Bestfoods France, APRIFEL.
References
Adlercreutz H, Fotsis T, Bannwart C, Wa¨ha¨la¨ K, Ma¨kela¨ T, Brunow G et al. (1986). Determination of urinary lignans and phytoestrogens metabolites, potential antiestrogens and anticarcinogens, in urine of women on various habitual diets. J Steroid Biochem 25, 791–797. Beer-Borst S, Hercberg S, Morabia A, Bernstein MS, Galan P, Galasso R et al. (2000). Dietary patterns in six European populations: results from EURALIM, a collaborative European data harmonization and information campaign. Eur J Clin Nutr 54, 253–262.
Belajova E, Suhaj M (2004). Determination of phenolic constituents in citrus juices: method of high performance liquid chromato-graphy. Food Chemistry 86, 339–343.
Bingham SA (1991). Limitations of the various methods for collecting dietary intake data. Ann Nutr Metab 35, 117–127. Bingham SA (2002). Biomarkers in nutritional epidemiology. Public
Health Nutr 5, 821–827.
Block G (1982). A review of validations of dietary assessment methods. Am J Epidemiol 115, 492–505.
Borriello SP, Setchell KDR, Axelson M, Lawson AM (1985). Produc-tion and metabolism of lignans by the human fecal flora. J Appl Bacteriol 58, 37–43.
de Vries JH, Hollman PC, Meyboom S, Buysman MN, Zock PL, van Staveren WA et al. (1998). Plasma concentrations and urinary excretion of the antioxidant flavonols quercetin and kaempferol as biomarkers for dietary intake. Am J Clin Nutr 68, 60–65. DuPont MS, Bennett RN, Mellon FA, Williamson G (2002).
Polyphenols from alcoholic apple cider are absorbed, metabolized and excreted by humans. J Nutr 132, 172–175.
Gonthier MP, Verny MA, Besson C, Remesy C, Scalbert A (2003). Chlorogenic acid bioavailability largely depends on its metabo-lism by the gut microflora in rats. J Nutr 133, 1853–1859. Hercberg S, Galan P, Preziosi P, Bertrais S, Mennen L, Malvy D et al.
(2004). The SU.VI.MAX Study: a randomized, placebo-controlled trial of the health effects of antioxidant vitamins and minerals. Arch Intern Med 164, 2335–2342.
Hertog MGL, Feskens EJM, Hollman PCH, Katan MB, Kromhout D (1993). Dietary antioxidant flavonoids and risk of coronary heart disease: the Zutphen Elderly Study. Lancet 342, 1007–1011. Hirvonen T, Pietinen P, Virtanen M, Ovaskainen ML, Hakkinen S,
Albanes D et al. (2001). Intake of flavonols and flavones and risk of coronary heart disease in male smokers. Epidemiology 12, 62–67. Hodgson JM, Chan SY, Puddey IB, Devine A, Wattanapenpaiboon N,
Wahlqvist ML et al. (2004). Phenolic acid metabolites as
biomarkers for tea- and coffee-derived polyphenol exposure in human subjects. Br J Nutr 91, 301–306.
Hodgson JM, Morton LW, Puddey IB, Beilin LJ, Croft KD (2000). Gallic acid metabolites are markers of black tea intake in humans. J Agric Food Chem 48, 2276–2280.
Horn-Ross PL, Barnes S, Kirk M, Coward L, Parsonnet J, Hiatt RA (1997). Urinary phytoestrogen levels in young women from a multiethnic population. Cancer Epidemiol Biomarkers Prev 6, 339–345.
Ito H, Gonthier M-P, Manach C, Morand C, Mennen L, Re´me´sy C et al. (2005). Polyphenol levels in human urine after intake of six different polyphenol-rich beverages. Br J Nutr 94, 500–509. Kaaks RJ (1997). Biochemical markers as additional measurements in
studies of the accuracy of dietary questionnaire measurements: conceptual issues. Am J Clin Nutr 65, 1232S–1239S.
Kirkman LM, Lampe JW, Campbell DR, Martini MC, Slavin JL (1995). Urinary lignan and isoflavonoid excretion in men and women consuming vegetable and soy diets. Nutr Cancer – Int J 24, 1–12. Knekt P, Jarvinen R, Reunanen A, Maatela J (1996). Flavonoid intake
and coronary mortality in Finland: a cohort study. Br Med J 312, 478–481.
Krogholm KS, Haraldsdottir J, Knuthsen P, Rasmussen SE (2004). Urinary total flavonoid excretion but not 4-pyridoxic acid or potassium can be used as a biomarker for the intake of fruits and vegetables. J Nutr 134, 445–451.
Lampe JW, Gustafson DR, Hutchins AM, Martini MC, Li S, Wa¨ha¨la¨ K et al. (1999). Urinary isoflavonoid and lignan excretion on a
Western diet: relation to soy, vegetable, and fruit intake. Cancer Epidemiol Biomarkers Prev 8, 699–707.
Manach C, Scalbert A, Morand C, Re´me´sy C, Jimenez L (2004). Polyphenols – food sources and bioavailability. Am J Clin Nutr 79, 727–747.
Manach C, Williamson G, Morand C, Scalbert A, Remesy C (2005). Bioavailability and bioefficacy of polyphenols in humans. I. Review of 97 bioavailability studies. Am J Clin Nutr 81, 230S–242S. Maskarinec G, Singh S, Meng L, Franke AA (1998). Dietary soy intake and urinary isoflavone excretion among women from a multiethnic population. Cancer Epidemiol Biomarkers Prev 7, 613–619.
Mennen L, Sapinho D, Ito H, Galan P, Hercberg S, Scalbert A (2006). Urinary flavonoids and phenolic acids as biomarkers of intake for polyphenol-rich foods. British Journal of Nutrition 96, 191–198. Nielsen SE, Freese R, Kleemola P, Mutanen M (2002). Flavonoids in
human urine as biomarkers for intake of fruits and vegetables. Cancer Epidemiol Biomarkers Prev 11, 459–466.
Noroozi M, Burns J, Crozier A, Kelly IE, Lean ME (2000). Prediction of dietary flavonol consumption from fasting plasma concentration or urinary excretion. Eur J Clin Nutr 54, 143–149.
Scalbert A, Manach C, Morand C, Re´me´sy C, Jime´nez L (2005). Dietary polyphenols and the prevention of diseases. Crit Rev Food Sci Nutr 45, 287–306.
Scalbert A, Williamson G (2000). Dietary intake and bioavailability of polyphenols. J Nutr 130, 2073S–2085S.
