Phytoestrogens Review - Food Standards Agency - Homepage

By Rachel Graham,2014-07-28 03:13
18 views 0
Phytoestrogens Review - Food Standards Agency - Homepage

5. Absorption, Distribution, Metabolism and

    Excretion of Phytoestrogens.


    5.1 The absorption, distribution, metabolism and excretion (ADME) of

    phytoestrogens have not been fully elucidated in human adults or infants. Most of the information concerns the isoflavones daidzein and genistein and to a lesser extent, the lignans enterodiol and enterolactone. This chapter reviews the ADME studies carried out in humans and, where indicated, provides relevant additional information gained from animal studies. There are no data available on the ADME of prenylated flavonoids or coumestans. Other gaps in the knowledge are highlighted in this chapter.



    5.2 Isoflavones are present in food mainly as glucosides. Although there is evidence to suggest that particular members of a related class of flavonoids are absorbed in their naturally occurring glucosidic forms (Hollman & Katan, 1997), this does not appear to be the case for the isoflavones. It is thought that isoflavones are absorbed as aglucones, which are more readily absorbed than the parent glucosides due to their higher hydrophobicity and lower molecular weight. Glucosides of isoflavones have not been identified in plasma. This is supported by recent evidence from Setchell et al (2002), which indicates that isoflavone glucosides are not absorbed intact across the enterocyte of healthy adults and shows that uptake requires hydrolysis of the isoflavone glucosides to their aglucone form. Absorption of aglucones takes place mainly in the small and large intestine (see Figure 5.1).Transfer

    5.3 It has been suggested that acid hydrolysis of glucosides occurs in the stomach (Kelly et al, 1993) although there is some disagreement about this (Piskula et al,

    1999). There is evidence that the liver and enterocytes of the human small intestine contain β-glucosidase enzymes capable of efficiently hydrolysing some, but not all, naturally occurring flavone and isoflavone glucosides (Day et al, 1998). β-

    Glucosidases associated with the gut microflora (including Lactobacilli,

    Bifidobacteria and Bacteroides) also play a role in glucoside hydrolysis (Xu et al,

    1995; Barnes et al, 1996). A study has also suggested that isoflavone glucosides can be converted to aglucones by enzymes in saliva (Allred et al, 2001). In humans, prior

    to absorption, the isoflavones may be further metabolised by the gut microflora, with genistein being converted to the hormonally inert p-ethyl-phenol and daidzein reduced

    to the oestrogenically active isoflavone equol and the non-oestrogenic O-

    demethylangolensin (O-DMA).

    5.4 Little information on the processing of lignans and coumestans prior to absorption has been reported. However, ingested lignans have been shown to undergo This is a draft report of the COT Working Group on Phytoestrogens which is open to consultation


    bacterial hydrolysis and metabolism. Colonic fermentation results in the removal of glucose residues, demethylation and dehydroxylation to the diphenol compounds, enterolactone and enterodiol, which are absorbed. Enterodiol may be further metabolised to enterolactone in the gut (Setchell & Adlercreutz, 1998; Kurzer & Xu, 1997) (see Figure 5.2).

    5.5Studies have suggested that isoflavones are more bioavailable in food if present as aglucones (as in fermented soy) than when present as glucosides (as in unprocessed soy). Hutchins et al (1995) reported that the recovery of urinary daidzein and genistein was higher when subjects consumed a diet consisting of tempeh (fermented soy) compared to a similar diet containing conjugated isoflavones in the form of unfermented soy pieces. In addition, Slavin et al (1998) reported that while

    fermentation decreased the isoflavone content of soy, the increased recovery of urinary isoflavones observed suggested fermentation (conversion to aglucones) increased the bioavailability of isoflavones. However, data from animal studies indicates that, although the initial rates of absorption and excretion for the aglucone forms were greater than for the conjugated forms, the total percentage recoveries in urine and faeces does not differ significantly. Thus, the extent of absorption and the bioavailability of aglucone and glucosidic forms are considered to be similar (King et

    al, 1996). A research project funded by the Food Standards Agency is currently investigating the effects of chemical form on the absorption of isoflavones in humans (FSA project T05010).


    1Bacterial activity in the human intestine

    5.6 Metabolism of isoflavones and lignans is mediated both by tissue enzymes and gut microflora either prior to absorption or during enterohepatic circulation (see paragraphs 5.49-5.59, Figure 5.1).

    5.7The importance of the gut microflora in the metabolism of phytoestrogens has been clearly established. Setchell et al (1984) showed that incubation of textured

    vegetable protein with cultured human faecal bacteria resulted in the formation of equol, a metabolic reduction product of daidzein (see Figure 5.3). Chang & Nair (1995) demonstrated the metabolism of daidzein to dihydrodaidzein, benzopyran-4,7-diol,3-(4-hydroxyphenol) and equol and of genistein to dihydrogenistein, when fermented with human faecal bacteria under anaerobic conditions. Antibiotic treatment in humans has been shown to inhibit the formation and excretion of enterolactone and enterodiol, which are bacterial metabolites of the lignans (Setchell et al, 1981).

    5.8Several groups of bacteria are known to possess β-glucosidase activity,

    including Lactobacilli, Bacteroides and Bifidobacteria (Hawkesworth et al, 1971).

    However, little research has been carried out to identify the types of bacteria specifically involved in the metabolism of isoflavones or lignans.

    1The reader is referred to the Institute of Environmental Health Report: Assessment of phytoestrogens in the diet (1997) for a detailed treatise on gut floral composition, inter-individual variations and implications for health.This is a draft report of the COT Working Group on Phytoestrogens which is open to consultation


    Figure 5.1 Schematic representation of absorption of daidzein from the gut. Prior to absorption from the gut, daidzin is converted to daidzein by gut microfloral enzymes. It is partially converted to glucuronide and sulphate conjugates by

    enzymes in the liver before entering the peripheral circulation. These conjugates can be excreted back into the gut from the liver via the bile duct

    (enterohepatic circulation) where they can be deconjugated by gut microfloral enzymes. They may then be re-absorbed or further transformed in the

    gut and absorbed.

    Gut wallOHO



    Deconjugation by Biliary OOOgut microfloraexcretionOHDaidzein-7-glucuronideHOOHOHOOHODaidzein-7-sulfate

    Peripheral circulationDaidzeinOHepatic conjugation to HOOSglucuronides and sulfatesOOOOHLiverOHODaidzein

    Further transformations (see figure 5.3)HOOAbsorption

    Faecal excretionUrinary excretion

    This is a draft report of the COT Working Group on Phytoestrogens which is open to consultation


    5.9 In humans, the upper third of the small intestine contains very low levels of bacteria, but this changes to a colon-like flora in the lower third (Heneghen, 1988). The distal part of the small intestine and the large intestine contain substantial 47numbers (10-10 bacteria/g wet weight) of Lactobacilli, Bacteroides and

    Bifodobacteria (Gorbach et al, 1967; Mitsuoka 1982) whereas the proximal end of the

    small intestine contains very few of these bacteria.

    5.10Some bacteria present in the large intestine also possess β-glucuronidase and

    arylsulfatase activity, which can liberate aglucones from conjugates excreted in the bile and render them available for reabsorption (Heneghen, 1988). Incubation studies with human faeces suggest that human intestinal bacteria from some, but not all, individuals can further metabolise and degrade isoflavones (Xu et al, 1995), thus

    preventing their reabsorption from the lower bowel. Consequently, the composition of intestinal microflora can have a profound effect both on the magnitude and pattern of isoflavone bioavailability.

    5.11Isoflavones that possess a 5-OH group, such as genistein, are much more susceptible to cleavage of the central ring system by rat intestinal bacteria (Griffiths & Smith, 1972). Whether the same is true with human faecal bacteria is unclear, although selective central ring cleavage of certain flavonoid compounds by certain strains of Clostridium isolated from human faecal flora has been shown (Winter et al,

    1989; Hur et al, 2002). Less faecal degradation should result in greater exposure to lignans and isoflavones in the circulation and increased urinary isoflavone excretion, although specific data are lacking.

    5.12Lignans have also been shown to produce enterodiol and enterolactone in humans via microfloral metabolism (Setchell & Adlercreutz, 1998) (see Figure 5.2). Conjugation

    5.13Once absorbed, isoflavones and lignans are efficiently reconjugated, either with glucuronic acid or, to a lesser extent, sulfate. In addition, some sulfoglucuronides may be formed. Conjugation takes place either in the liver with hepatic UDP-glucuronosyl transferase or sulfotransferase enzymes (Knight & Eden, 1996; Bingham et al, 1998; Setchell, 1998), or within the intestinal epithelium, which has also been shown to possess glucuronosyl transferase and sulfotransferase activity (Sfakianos et

    al, 1997). As a consequence, isoflavones and lignans are present in the circulation in predominantly conjugated forms.

    5.14Zhang et al (1999) have shown that rat microsomal UDP-glucuronosyl

    transferase has a greater affinity for genistein than daidzein in vitro. However, it

    remains unclear as to the relevance of this finding to humans.

    This is a draft report of the COT Working Group on Phytoestrogens which is open to consultation


    Figure 5.2 Example conversion of lignans to enterolactone and enterodiol by

    human faecal flora.

    For example, secoisolariciresinol monoglucoside is metabolised to enterodiol through hydrolysis of the

    sugar moiety, dehydroxylation, and demethylation. Enterodiol can then be further

    converted to enterolactone. Matairesinol is converted to enterolactone by gut bacteria

    through dehydroxylation and demethylation (Adapted from Kurzer & Xu, 1997).



    Secoisolariciresinol Matairesinolmonoglucoside








    Further metabolism


    5.15The metabolism of unconjugated isoflavones is complex (see Figure 5.3). Studies have revealed several diphenolic metabolites can be produced during the intermediary metabolism of daidzein and genistein. The intermediate metabolites can be produced either from microfloral or liver metabolism and include 6'-OH-O-

    desmethylangolensin (6’OH-DMA), dihydrogenistein, O-demethylangolensin (O-

    DMA), di and tetrahydrodaidzein and equol (Adlercreutz et al, 1987; Kelly et al,

    1993; Joannou et al, 1995; Yasuda & Ohsawa, 1998). It has been shown that some aglucones may undergo metabolism mediated by cytochrome P450 (CYP) enzymes (Roberts-Kirchhoff et al, 1999). Incubations with genistein in the presence of human 22recombinant CYP1A1, 1A2, 1B1, or 2E1 isoforms resulted in the formation of one predominant and two minor unidentified metabolites. Incubation with CYP3A4 catalysed the formation of two unidentified products. All metabolites were considered to be hydroxylation products. Similarly, transformation of the flavonoid galangin to kaempferol then to quercetin in rat liver microsomes is thought to be CYP-mediated, with the latter step being attributed specifically to CYP1A1 (Silva et

    al, 1997). Additionally, metabolism studies with cultured (T47D breast cancer) cells 22CYP1A1 and CYP1B1 activities are low or absent in human liver2

    This is a draft report of the COT Working Group on Phytoestrogens which is open to consultation


have shown that biochanin A and genistein undergo methylation and hydroxylation

    reactions as well as sulfate ester formation (Peterson et al, 1998).

    35.16Following incubation of multiply labelled H-genistein with rat caecal and

    faecal or human faecal cultures, the genistein metabolites: dihydrogenistein, 6’-OH-O-DMA and were detected by radio-HPLC. Further hydrolysis by the gut microflora of human and rat resulted in the formation of 4-hydroxyphenol-2-propionic acid (Coldham et al, 2002). Previous work has suggested that 4-p-ethylphenol was the

    final metabolite in the plasma and urine of rats and humans after administration of genistein (Barnes et al, 1998; King, 1998; Setchell, 1998). However, unlike the study by Coldham et al (2002) these studies were conducted with unlabelled genistein. 5.17Five metabolites of genistein were identified in rats following an oral dose of 4 mg genistein/kg bw. The metabolites were identified as α-genistein glucuronide,

    dihydrogenistein glucuronide, genistein sulphate, dihydro-genistein and 4-hydroxyphenyl-2-propionic acid (Coldham et al, 1999). The last of these metabolites

    is thought to be a microfloral product of dihydrogenistein. The major metabolites of genistein identified in the rat are therefore the same as those identified in man.Figure 5.3 Proposed phase I metabolism of daidzein and genistein based on

    human urinary metabolites.

    Adapted from Joannou et al, 1995.






    EquolHOOCH3CHHOOH3O-demethylangolensin OHHO(O-DMA)6'-Hydroxy-O-demtheylangolensin (6'-OH-DMA)


    5.18A range of lignan metabolites have been identified in human urine: enterolactone, enterodiol, matairesinol, lariciresinol, isolariciresinol, secoisolariciresinol and the tentatively identified 7α-OH matairesinol and 7’-OH

    enterolactone (Adlercreutz et al, 1995).

    This is a draft report of the COT Working Group on Phytoestrogens which is open to consultation


    5.19Oxidative metabolites of enterodiol and enterolactone have been identified in the urine and bile of female rats (Niemeyer et al, 2000). Following intraduodenal

    administration of enterodiol or enterolactone to bile duct-catheterised female rats, 5 hydroxylated biliary metabolites were produced from enterodiol and 11 were produced from enterolactone. The precise identity of these compounds was not determined.


    5.20Isoflavone and lignan phytoestrogens have been detected in a number of body fluids such as urine, plasma, faeces, prostatic fluid, semen, bile, saliva, breast milk, breast aspirate and cyst fluid. The major isoflavones and their metabolites detected in the blood and urine of humans and animals are daidzein, genistein, equol and O-DMA

    (Adlercreutz et al, 1995; Knight & Eden, 1996). Lignans identified in human plasma and urine include enterolactone, enterodiol, lariciresinol and isolariciresinol (Adlercreutz et al, 1987; Jacob et al, 2002).

    5.21In rats, relatively high levels of daidzein in the plasma, liver, lung and kidney were observed 15 min after intravenous injection. Lower levels were found in skeletal muscle, spleen, heart, testis and brain (Yueh & Chu, 1997).

    5.22It has been shown that isoflavones can cross the blood brain barrier in rats (Setchell, 1998). Following intraperitoneal administration to rats, genistein rapidly appears in brain and then in microdialysate fluid from the corpus striata together with 14its metabolite p-ethyl-phenol. Following oral administration of C-genistein to rats,

    the levels in reproductive organs (vagina, uterus, ovary, and prostate) were higher than in other peripheral organs. The major plasma-derived compounds were genistein glucuronides and 4-hydroxyphenyl-2-propionic acid with only trace amounts of parent compound (Coldham & Sauer, 2000).

    5.23The concentrations of genistein in plasma and selected tissues in weanling (plasma only) and adult rats exposed to genistein in utero, through maternal milk and

    via the diet (5, 100 or 500 mg genistein/kg diet) were measured (Chang et al, 2000).

    Plasma and tissue concentrations of genistein increased dose dependently. Genistein was predominantly (95-99%) present in conjugated form in plasma but to a much lesser extent in tissues, unconjugated genistein ranged from 18-100% (see Table 5.1). Gender differences in the tissue concentrations of genistein and the proportion of genistein in conjugated form were evident particularly in the liver, thyroid and mammary glands. Little genistein was detected in the brain.

    5.24Holder et al (1999) demonstrated that genistein was predominantly conjugated (90%) to glucuronic acid with minor quantities of sulfate conjugates in the plasma of rats. Janning et al (2000) also demonstrated that following a single dose of daidzein (100 mg/kg bw) administered to adult female rats, daidzein was detected in plasma almost exclusively in its conjugated form.

    This is a draft report of the COT Working Group on Phytoestrogens which is open to consultation


    Table 5.1 Plasma and tissue concentrations of genistein in male and female rats. Concentrations and percentage unconjugated genistein shown are from animals (n= 6) exposed to 500mg genistein/kg diet. Adapted from Chang et al (2000).

    TissueGenistein Concentration pmol/mg (% aglucone)


    aPlasma (adult)6 µmol/L (<5%)7.9 µmol/L (<5%)aPlasma (weanling)1.9 µmol/L (<5%)2.1 µmol/L (<5%)

    Mammary glands0.8 (24%)2.4 (49%)

    Thyroid0.4 (25%)1.2 (18%)

    Liver0.7 (34%)7.33 (77%)bBrainLodLod

    Prostate1.1 (45%)

    Testes0.6 (11%)

    Ovary1.1 (80%)

    Uterus1.4 (100%)a Concentrations in plasma are given in µmol/L.b Concentrations in brain were at or below the limit of detection (0.5 pmol/mg).


    5.25Many studies have examined the excretion profiles of phytoestrogens. However, few have examined the pharmacokinetics of these compounds in plasma. Isoflavones

    5.26The most complete pharmacokinetic study of isoflavones in humans, published to date, is that by Watanabe et al (1998). Seven adult males (previously

    receiving an otherwise low isoflavone diets for 6-days) received a single dose of baked soybean powder containing 103 µmol (26 mg) daidzein and 112 µmol (30 mg)

    genistein for 10 days. The isoflavone concentrations in plasma, urine and faeces are shown in Table 5.2. Plasma genistein concentrations had significantly increased 2 hours after ingestion and reached maximum concentrations by 8 hours. The plasma concentration of daidzein peaked at approximately the same time but was lower than genistein. Plasma levels of O-DMA and equol peaked later than genistein and

    daidzein in 2/7 and 4/7 subjects, respectively.

    5.27In contrast to plasma, daidzein was the major compound present in urine. Recoveries of ingested daidzein and genistein in urine were 35.8% and 17.6%, respectively. Equol was detectable in the urine of 2/7 subjects and was observed after the peak of daidzein excretion.

    This is a draft report of the COT Working Group on Phytoestrogens which is open to consultation


    Table 5.2 Total recovery of isoflavones, O-DMA and equol in urine and faeces of

    men for 3 days after ingestion of 60g kinako (baked soybean powder). Data adapted from Watanabe et al, 1998.

    SubjectDaidzein (µmol)O-DMA (µmol)Equol (µmol)Genistein (µmol)











    Recovery aaaa3644272182(%)aThese values are the percentage of recovery of daidzein ingested. DMA = O-desmethylangolensin.

    5.28Most of the isoflavones were recovered in the faeces 2-3 days after ingestion. Excretion of total diphenolics, including equol, was often higher in the second and third days than in the 24 hours after ingestion. This suggests that faecal isoflavones and equol were derived from biliary excretion. Total recovery of daidzein and its metabolites, O-DMA and equol, from urine and faeces was ~55% whereas ~20% of administered genistein was recovered unchanged. Plasma half-lives for genistein and daidzein were 8.4 and 5.8 hours, respectively.

    5.29Significant inter-individual variation was observed in the plasma and urinary levels of equol and O-DMA. In addition, a biphasic peak in plasma and urine

    genistein and daidzein concentrations was observed in a number of subjects, suggesting enterohepatic circulation.

    5.30A recently completed FSA funded study investigated the pharmacokinetic 13profile of C-labelled daidzein and genistein administered to pre-menopausal women (n=20). Four separate experiments were conducted. In the first, subjects received a single oral dose of 0.4 mg/kg bw of each compound. This was repeated in the second experiment to assess intra-individual variability in metabolism. To assess the effects of dose on pharmacokinetics 0.8 mg/kg bw doses were administered in the third experiment and in the fourth 0.4 mg/kg bw doses were administered following ingestion of 50 mg/kg bw isoflavones in food for 7 days (FSA project report T05019).5.31No significant differences in serum and urinary profiles were observed between repeat doses demonstrating little intra-individual variability in metabolism of daidzein and genistein. Mean plasma half-lives for both compounds were 7.7 hours. Peak plasma concentrations were greater for genistein than daidzein at the doses used suggesting genistein is the more bioavailable isoflavone. The bioavailability of both compounds, as assessed by AUC (area under the curve) increased with dose. Both compounds had a large volume of distribution. The mean recoveries of daidzein and This is a draft report of the COT Working Group on Phytoestrogens which is open to consultation


    genistein were 30% and 9%, respectively indicating significant metabolism to other compounds (FSA project report T05019).

    5.32The plasma and urine concentrations of daidzein and genistein peaked 7 to 8 hours after ingestion of a single soybean based meal by adult men (n= 6) (King & Bursill, 1998). Elimination half-lives were approximately 5 and 6 hours for daidzein and genistein, respectively. Although the urinary excretion of daidzein was greater than that for genistein, the ratios of the plasma AUCs for the respective isoflavones were similar to the ratio of concentrations present in the soybean meal, indicating similar bioavailabilities for the two isoflavones (King & Bursill, 1998). Similar findings were reported by Xu et al (2000). These studies apparently conflict with

    those of Watanabe et al (1998) and the FSA study (FSA final report T05019).

    However, a study by Setchell et al (1998) employing pure isoflavones administered as

    a single bolus dose, has also demonstrated that the apparent bioavailability of daidzein and genistein (determined from plasma appearance and disappearance curves), are similar. Peak plasma concentrations were usually attained between 6-8 hours after ingestion and plasma half-lives were approximately 8 hours.

    5.33In a study reported by Xu et al (1994), the plasma concentrations of daidzein

    and genistein in women (n= 12) were both significantly increased 6.5 hours after the consumption of a soy milk powder drink (containing a daidzein:genistein ratio of 44:56). Faecal excretion was low (1-2% of isoflavones ingested). On the basis that urinary excretion of daidzein (21%) was much higher than that of genistein (9%), the authors concluded that daidzein had a greater bioavailability.

    5.34Coward et al (1996) reported that the ratios of genistein to daidzein

    concentrations in plasma did not vary significantly from that present in a soy beverage, thus suggesting bioavailability of the two isoflavones in humans was similar. However, daidzein conjugates were reported to be more bioavailable than genistein conjugates when administered to rats as a soy extract (King, 1998). 5.35The pharmacokinetics of genistein and daidzein was compared with that of their glucosides in premenopausal women (n= 19) fed a 50 mg single bolus dose of each compound (Setchell et al, 2001). All compounds were efficiently absorbed from

    the intestinal tract however, the glucosides took longer to achieve maximum plasma concentrations (9.3 and 9.0 h compared with 5.2 and 6.6 h for genistein and daidzein, respectively). The bioavailability of the isoflavones was greater when ingested in the glucoside form.

    5.36Gender differences in bioavailability were reported in rats (Coldham & Sauer, 14C-genistein, after an oral dose of 4 mg/kg bw for 2000). The mean total excretion of

    male and female rats, was approximately 67 and 33 % in faeces and urine, respectively within 168 hours of dosing. Mean and maximal concentrations in the plasma were higher in male than in female with half-lives of 12.4 hours and 8.5 hours, respectively.

    5.37In a study by Janning et al (2000), the plasma concentration-time curve

    following intravenous administration of daidzein to female rats could be fitted to a triexponential model with a final half-life of 4 hours. The oral bioavailability was 9.7% and 2.2% when 10 and 100 mg daidzein/kg bw, respectively were administered, This is a draft report of the COT Working Group on Phytoestrogens which is open to consultation


Report this document

For any questions or suggestions please email