Phytoestrogens are a class of heterocyclic polyphenols existing in plants whose composition includes two hydroxyl groups and one phenolic ring (Figure 1). The phenolic ring controls how tightly phytoestrogens bind to receptors. Because phytoestrogens function as both estrogen agonists and antagonists, they are often referred to as selective estrogen receptor modulators (SERMs) (Brzezinski and Debi, 1999). As an estrogen agonist, phytoestrogens can combine with estrogen receptor (ER) to play a weak estrogen effect (Rietjens et al., 2017). As estrogen antagonists, they can block estrogen receptors and inhibit estrogen activity, causing anti-estrogen effect. The bioavailability of phytoestrogens depends on the form of action, dosage, individual metabolism, other drug intake factors, target tissue concentration dependence, and whether endogenous estrogen exists or not (Joannou et al., 1995; Xu et al., 1995; Wiseman, 1999; Glazier and Bowman, 2001).

The reported phytoestrogens include isoflavones, coumarins, lignans, anthracenes, chalcones, and saponins (Matsuda et al., 2001; Chan et al., 2002; Rietjens et al., 2017). According to the structure of phytoestrogens, they are mainly divided into flavonoids, stilbenes, lignans, coumarins, and mycoestrogens. (i) Flavonoids, such as daidzein, genistein, calycosin, and so on, mainly exist in Leguminosae. They usually exist in the conjugate forms of genistein, daidzein, puerarin, daidzein, sandalwoodin, and sissotrin (Rietjens et al., 2017). (ii) Stilbenes, such as 6-isoprenyl naringenin, 6-vanillyl naringenin, 8-isoprenyl naringenin and isoflavones, 8-isoprenyl naringenin is the phytoestrogen with the most obvious estrogenic effect. They mainly exist in higher plants such as pine, mulberry, Gnetaceae, Cyperaceae, Fabaceae, Dipterocarpaceae, and Vitaceae (Tanwar et al., 2021). (iii) Lignans, including intestinal diols and enterolactones, are converted from lignan precursors under the action of intestinal flora (Lampe, 2003). Other intestinal lipid precursors identified include Arctigenin, 7-hydroxy lycopene, lariciresinol, pinoresinol and syringaresinol (Meagher et al., 1999; Heinonen et al., 2001). Lignans are abundant in flaxseed, whole wheat bread, fruits, vegetables, sesame, tea and other foods (Rietjens et al., 2017; Tanwar et al., 2021). (iv) Coumarins include psoralen, coumarins, 4' -methoxycoumarin, angelica visfatin, repensol, trifoliate phenol, etc. They mainly exist in legumes, especially in edible plants, such as peas, mung bean sprouts, alfalfa, clover sprouts (Rietjens et al., 2017). (v) Mycoestrogens, a natural estrogen produced by fungi, is harmful to animals when eating contaminated feed. Zearalenone (ZEA) is a fungal estrogen that has been studied extensively. ZEA is widely found in contaminated foods. Due to its strong estrogenic activity, ZEA is considered to be a reason for female reproductive changes (Xu et al., 2018).

Liver lipid metabolism includes lipid uptake and production, output and oxidation. Breaking one or more balance can promotes liver steatosis (Jones, 2016; Ipsen et al., 2018). In addition, postmenopausal women are prone to systemic fat redistribution due to estrogen deficiency. The risk of abdominal obesity in postmenopausal women is significantly higher than that in premenopausal women (5 times) (Donato et al., 2006). Therefore, restoring the lipid metabolism balance is a crucial link in the prevention and treatment of postmenopausal NAFLD (Figure 2).

Genistein is the most abundant phytoestrogen in soybean and is one of the most studied phytoestrogens. Several studies have found that genstein can reduce the body weight, liver weight, serum and liver lipid levels (triglyceride, cholesterol, serum free fatty acids) of ApoE (−/−) (Jeon et al., 2014), C57BL6J mice (Kim et al., 2010), Sprague-Dawley (SD) rats (Liu H. et al., 2017) and SD-ovariectomized rats (Witayavanitkul et al., 2020) by the following mechanisms: (i) Reducing free fatty acid (FFA) intake and liver lipid source (Jeon et al., 2014). Genstein reduced hepatic TG and FFA by inhibiting scavenger receptor, CD36 and scavenger receptor A uptake by oxidized low density lipoprotein. (ii) Inhibition of lipid synthesis related gene expression (Liu H. et al., 2017; Seidemann et al., 2021). On the one hand, genstein could downregulate the expression of AMP-activated protein kinase/acetyl-coenzyme A carboxylase (AMPK/ACC) signaling pathway, inhibited the expression of downstream fatty acid synthase and 3-phosphate glyceryl transferase (GPAT), and reduced liver lipid synthesis. On the other hand, Genstein inhibited the expression of sterol regulatory element binding protein 1 (SREBP1), thereby reducing the hepatic de novo adipogenesis (DNL) pathway. (iii) Reducing lipid peroxidation and promoting gene expression related to FFA oxidative decomposition (Kim et al., 2010; Zhong et al., 2017; Seidemann et al., 2021). Genstein promoted the expression of peroxisome proliferator-activated receptor α (PPARα), carnitine palmitoyl transferase-1 (CPT-1) and acyl-CoA oxidase (ACO), thereby accelerating the β-oxidation of FFA, and ultimately reducing liver lipid accumulation and lipid toxicity. Amanat et al. (Amanat et al., 2018) randomly divided 82 NAFLD subjects into treatment group and placebo group. The treatment group was supplemented with 250 mg genistein daily for 8 weeks. At the end of the experiment, it was found that the waist-to-hip ratio, body fat percentage and TG of NAFLD subjects supplemented with genistein were significantly lower than those of the placebo group. Another cross-sectional study of 6786 adults in China (Wang et al., 2022) found that dietary isoflavone intake was negatively correlated with the prevalence of NAFLD, hyperlipidemia and hypertension.

In Ji-Hye Jung study (Jung and Kim, 2013), mice were fed with a high cholesterol/fat diet (HCD) and treated with different concentrations of black soybean powder, showing different effects on improving liver and adipose tissue lipid metabolism. The levels of total cholesterol (TC) and TG and the expression of SREBP2 in the liver of mice treated with 4% black soybean powder were significantly lower than those in the model group. Black soybean powder could stimulate the secretion of adiponectin, activate the expression of pAMPK and eliminate FFA in the liver. Similarly, daidzein treatment (Cao et al., 2013) increased serum adiponectin levels and decreased body weight gain, visceral fat gain and HOMA-IR index in OVX rats. Naoki Nanashima et al. (Nanashima et al., 2020) used 3% blackcurrant to treat OVX rats for 3 months. At the end of the experiment, it was found that 3% blackcurrant significantly reduced the body weight, visceral fat weight, TG, TC and low-density lipoprotein (LDL) of rats. More importantly, hematoxylin and eosin staining (HE) showed that blackcurrant (Ribes nigrum L.) extract (BCE) reduced the diameter of adipocytes and the score of NAFLD activity. C57BL/6J mice fed with genistein (Kim et al., 2010) showed a dose-dependent decrease in body weight and lipid levels. Genistein also inhibited adipocyte hypertrophy and adipogenesis by down-regulating the expression of LXRα, SREBP1c and PPARγ. In Narrerat study (Sutjarit et al., 2018), it also was found that comosa Roxb could inhibit adipocyte size and regulate adipokine secretion in OVX rats. Panneerselvam et al. (2016)found that soy isoflavones (150 mg/kg body weight/day, 8 weeks) improved hepatic steatosis in HFD-fed ovariectomized Wistar rats by down-regulating insulin-inducible gene 2 (insig2) and PPARα expression in adipose tissue, inhibiting adipocyte differentiation and reducing fat formation. Kim et al. (2020) found that coumarin (100 μg/20g body weight/day, 2 weeks) can activate brown adipose tissue (BAT) in HFD-fed C57BL/6J mice, increase BAT mitochondria, and accelerate BAT energy metabolism. The Shen et al. (2019); Kang et al. (2020)have proved that Genstein and Secoisolariciresinol diglucoside (SDG) can upregulate the expression of peroxisome proliferator-activated receptor gamma coactivator 1α (PGC1α), promote white fat browning, and increase body heat production and reduce fat accumulation through AMPK pathway. These basic studies provide evidence for phytoestrogens to prevent lipid metabolism disorders and NAFLD in postmenopausal women. Table 1 lists other studies on phytoestrogens improving lipid metabolism in ovariectomized models (Dai et al., 2004; Böttner et al., 2008; Pakalapati et al., 2009; Miller et al., 2015; Cross et al., 2017; Chen et al., 2018; Zheng et al., 2018; Zingue et al., 2018; Huang et al., 2020; Tang et al., 2021).

Insulin resistance (IR) is a decline in insulin sensitivity or responsiveness in target organs (such as liver, adipose tissue, skeletal muscle, etc.) (Lebovitz, 2001). IR is the key link in the occurrence of NAFLD. As the initiating factor of “first hit”, IR is closely related to “second hit” factors such as lipid peroxidation, oxidative stress and inflammatory response (Day, 2002; Fang et al., 2018; Bessone et al., 2019). Due to estrogen deficiency in postmenopausal women, the expression of insulin receptor and insulin receptor substrate-1/2 (IRS-1/2) is reduced, and the insulin signal transduction is weakened, resulting in IR and eventually inducing NAFLD (Mauvais-Jarvis et al., 2017).

Genistein could reduce the serum insulin level in NASH model of SD rats and improve HOMA-IR in a dose-dependent manner (Yin et al., 2019). Similarly, in the OVX model of SD rats, supplementation of daidzein (50 mg/kg body weight/day, 12 weeks) could also reduce body weight, HOMA-IR and fasting insulin level (Cao et al., 2013). Another study showed that genistein alone or in combination with metformin significantly reduced fasting blood glucose (FBS) in HFD mice by reducing glucose 6-phosphatase (G6Pase) and increasing glycogen synthase kinase 3β (GSK-3β) phosphorylation, thereby inhibiting gluconeogenesis (Zamani-Garmsiri et al., 2021). In vitro studies by Tomasz et al. (Charytoniuk et al., 2019) found that enterolactone could reduce the phosphorylation levels of Serine/threonine kinases (AKT) and AMPK in HepG2 cells induced by palmitic acid, and ultimately improve liver insulin sensitivity.

In addition, clinical studies have yielded encouraging findings. In a randomized double-blind controlled trial (Amanat et al., 2018), compared with placebo, genistein supplementation (250 mg/day, 8 weeks) could reduce insulin levels and HOMA-IR in NAFLD patients. In another randomized double-blind control experiment involving 54 postmenopausal patients with type 2 diabetes, compared with placebo, genistein supplementation (108 mg/day, 12 weeks) could significantly reduce the subjects’ FBS, glycosylated hemoglobin, TG and malondialdehyde (MDA) (Braxas et al., 2019). However, the dose and time of genistein used in these two studies are different, which may be related to different dosage forms and subjects of genistein. A meta-analysis (Liu Y. et al., 2017) concluded that genistein significantly improved blood glucose levels and insulin sensitivity in postmenopausal women and that long-term treatment may be more effective than short-term use. Therefore, more clinical trials are still needed to determine the optimal dose and course of treatment of phytoestrogens for different diseases.

Oxidative stress occurs when the production of oxidative molecules (such as superoxide, hydrogen peroxide) exceeds the scavenging capacity of antioxidant molecules (such as catalase (CAT), glutathione peroxidase (GSH-Px), superoxide dismutase (SOD)) (Sies, 2015). When oxidation and antioxidant imbalance occurs, it will cause lipid and protein peroxidation, nucleic acid oxidative modification, and other metabolic diseases such as NAFLD.

Phytoestrogens have been proved to have antioxidant effects in various models. In OVX-NASH rat model, compared with OVX model group, genistein supplementation could significantly reduce liver MDA and increase reduced glutathione (GSH) levels (Witayavanitkul et al., 2020). In OVX-LDLR (−/−) atherosclerosis model mice and HAEC cell model, dioscin could reduce the levels of MDA and reactive oxygen species (ROS), and increase the levels of GSH and nicotinamide adenine dinucleotide phosphate oxidase 4 (NOX4), which was related to the activation of PGC-1α/ERα pathway (Yang et al., 2019). In the aging rat model, compared with the control group, the supplementation of Fructus Corni extract increased liver SOD, CAT and GPX, and decreased liver lipid peroxidation, suggesting that Fructus Corni extract may be a ROS scavenger (Hamden et al., 2009). Similarly, in the Granulosa cells model, genistein significantly increased mitochondrial membrane potential and enhanced the expression of SOD, GPX, CAT and adenosine-3′, 5′-cyclic monophosphate (cAMP), which was related to cAMP-PKA signaling pathway (Luo et al., 2020). In clinical studies, it is also found that phytoestrogens have antioxidant effects. Compared with placebo, genistein supplementation could reduce the serum MDA levels in NAFLD patients and postmenopausal type 2 diabetes (T2DM) patients (Amanat et al., 2018; Braxas et al., 2019). In summary, phytoestrogens maintain the balance of oxidation/antioxidant system directly or indirectly to achieve antioxidant effect in vivo and in vitro models. Table 2 summarizes the basic research of other phytoestrogens on improving oxidative stress in ovariectomized models (Baeza et al., 2010; Chulikhit et al., 2021; Jdidi et al., 2021; Ltaif et al., 2021).

Estrogen at physiological concentration can inhibit the release of tumor necrosis factor α (TNFα), interleukin-6 (IL-6) and interleukin-1β (IL-1β). A cross-sectional study (Rodrigues et al., 2014) found that serum IL-6 and TNFα levels in postmenopausal NAFLD patients with metabolic syndrome were higher than those in the control group, suggesting that estrogen deficiency would promote or aggravate the development of NAFLD. Phytoestrogens have a good inhibitory effect on liver inflammation in NAFLD. In patients with NAFLD, Amanat’s study (Amanat et al., 2018) demonstrated that genistein 250 mg daily for 8 weeks could lower TNFα and IL-6 levels and alleviate liver inflammation.

A study from author Cao YK (Cao et al., 2013) has shown that supplementation of daidzein (50 mg/kg) could reduce serum inflammatory factors in OVX rats, such as TNFα, IL-6. Another study from author Nanashima N (Nanashima et al., 2020) showed that dietary supplementation of 3% Ribes nigrum could reduce TNFα, IL-6 and IL-1β in OVX rats and improve the pathological state of liver inflammation. Similarly, the supplementation of 500 mg/kg flavonoids quercetin (Quercitrin) could increase the serum estrogen level in OVX mice with NAFLD and reduce the expressions of TNFα, IL-6 and IL-1β (Hur et al., 2020). In the studies about the beneficial effects of genistein in the C57BL/6 mouse NASH model, Zamani-GF et al.and Gan M et al.found the anti-inflammatory pathway of genistein. On the one hand, genistein promoted the transformation of macrophages into M2-type anti-inflammatory phenotype, reduced the infiltration of M1-type pro-inflammatory macrophages, thereby reduced the secretion of pro-inflammatory factors and ultimately suppressed expression of NF-κB (Zamani-Garmsiri et al., 2021). On the other hand, genistein directly inhibited the secretion of inflammatory factors and achieved anti-inflammatory effect by up-regulating the expression of miR-451 in liver (Gan et al., 2019). Many other studies have confirmed that genistein exerts anti-inflammatory effects by inhibiting the expression of Toll-like receptor 4 (TLR4) and reducing the levels of downstream TNFα, IL-6, endotoxin and 8—isoprostaglandin (Yalniz et al., 2007; Ji et al., 2011; Incir et al., 2016; Yin et al., 2019). Xu et al. (Xu et al., 2021) found that genistein regulated (CD68 + CD163)/(CD68 + CD206) protein expression through JAK2/STAT3/SOCS3 signaling pathway, thereby regulating the proportion of liver M1/M2 macrophages, reducing the level of IL-1β, IL-6, TNF-α and monocyte chemotactic protein 1 (MCP-1), and playing an anti-inflammatory role. (Figure 3).

With the development of ‘gut-liver axis’ theory and gene sequencing technology, the role of intestinal flora in the pathogenesis of NAFLD has received extensive attention, which provides a new target for the treatment of NAFLD. Clinical studies (Zhao et al., 2019) have shown that compared with premenopausal women, the ratio of M. intestinalis/Bacteroides (F/B) in postmenopausal women increases, and higher F/B is related to obesity. This may be due to decreased estrogen levels in postmenopausal women causing fat redistribution, mainly abdominal fat increase. Therefore, maintaining intestinal microecological balance in postmenopausal women may become a new strategy for the treatment of postmenopausal NAFLD.

There is an interaction between phytoestrogens and intestinal flora. Phytoestrogens are mainly converted into compounds with estrogenic activity by intestinal microorganisms in vivo. For example, intestinal microorganisms can convert daidzein and genistein into equol derivatives (Zhao et al., 2019; Á et al., 2020), so the composition of intestinal flora will affect the metabolism of phytoestrogens. On the other hand, studies have confirmed that phytoestrogens and their metabolites can also regulate and reshape the composition of intestinal microorganisms. A study (Naudhani et al., 2021) found that formononetin regulated intestinal microbial balance by increasing the number of Clostridium aldenense, Clostridaceae unclassified, cluster true bacteria, acetate and butyrate-producing bacteria, and maintained the integrity of the intestinal membrane by up-regulating the expression of Muc-2 and occludin. Consistent with previous studies, another study (Cardona et al., 2013) found that genistein reduced serum endotoxin levels by selectively increasing Akkermansia muciniphila to alter intestinal microflora in mice, thereby achieving anti-obese effects. In addition, berberine could increase the abundance of beneficial bacteria in OVX rats, such as Bacteroides, Bifidobacterium, Lactobacillus and Akmania (Fang et al., 2021). Moreover, genistein, daidzein and Humulus lupulus L. extract could improve intestinal mucosal barrier function, increase intestinal microbial diversity and reduce the abundance of pathogenic bacteria (Hamm et al., 2019; Ou et al., 2019; Ortega-Santos et al., 2020). A clinical study (Ou et al., 2019) found that compared with placebo, genistein (50 mg/day) treatment for 2 months could increase the number of intestinal Verrucomicrobia in obese patients, reduce serum endotoxin level and improve IR.

Ko SH’s study found that postmenopausal women had a higher incidence of NAFLD than premenopausal women (Ko and Kim, 2020). Importantly, studies found that the longer estrogen deficiency, the higher the risk of liver fibrosis in postmenopausal NAFLD patients (Klair et al., 2016). A number of studies have found that estrogen can inhibit the activation of hepatic stellate cells and prevent the progression of fibrosis. (Li et al., 2021; Lin et al., 2021; Šisl et al., 2022).

In vivo and in vitro studies have confirmed that phytoestrogens have antifibrotic effects. Many studies have confirmed that calycosin can improve the C57BL/6 liver fibrosis mouse model induced by carbon tetrachloride (CCL4). Studies by Zhang et al. (Zhang et al., 2021) and Deng et al. (Deng et al., 2018) confirmed that calycosin could improve liver fibrosis in C57BL/6 mice induced by CCL4. Zhang et al.and Deng et al.confirmed that calycosin can improve CCL4-induced liver fibrosis in C57BL/6 mice.

They reported a variety of mechanisms. Firstly, calycosin could increase matrix metalloproteinase-1 (MMP-1) expression, inhibit tissue inhibitors of metalloproteinases-1 (TIMP-1) expression, increase MMP-1/TIMP-1 ratio, inhibit collagen synthesis, and balance MMP-1/TIMP-1 system (Zhang et al., 2021). Secondly, calycosin inhibited fibrosis by increasing ERβ expression and activating JAK2-STAT3 pathway. Thirdly, calycosin significantly inhibited the proliferation and migration of activated hepatic stellate cells (HSCs) (Deng et al., 2018). Ganai et al. (Ganai and Husain, 2017) induced liver fibrosis in rats with d-galactosamine (D-GalN) and supplemented them with genistein (5 mg/kg body weight) for 12 weeks. At the end of the experiment, it was found that genistein could inhibit the accumulation of α smooth muscle actin (αSMA), which is a marker of HSC cell activation. TGF-β/Smad signaling pathway is a star pathway in the process of liver fibrosis. Ganai et al. (Ganai and Husain, 2017) found that genistein could play an anti-fibrosis role by blocking TGF-β/Smad signaling pathway. Xu et al. (Xu et al., 2021) confirmed through in vivo experiments that the supplementation of genistein also improved liver fibrosis in rats induced by dimethylnitrosamine (DMN). Genistein could inhibit the expression of αSMA and type I collagen α1 in rat liver and improve liver pathological injury. At the same time, in vitro experiments performed with genistein on HSC cell line LX2 cells confirmed that genistein could inhibit the viability and proliferation of LX2 cells, and it was important to induce LX2 cell cycle arrest in G0/G1 phase. Although Ganai et al. and Xu et al. used different methods to induce liver fibrosis, they all clarified the anti-hepatic fibrosis effect of genistein.