In order to explore the potential use of natural medicinal plants and plant extracts for the treatment of NAFLD, NASH and metabolic-associated fatty liver disease, a comprehensive search was conducted using relevant key words, such as ‘medicinal plants’, ‘plant extracts’, ‘non-alcoholic fatty liver disease’ and ‘non-alcoholic steatohepatitis’. The search was performed across various databases, such as Scopus, PubMed, Springer, NCBI, Google Scholar, ScienceDirect and Web of Science. For this search, studies conducted between January, 2016 and November, 2023 were considered, with a focus on in vivo studies that evaluated the effectiveness of natural medicinal plants for the treatment of NAFLD. The mechanisms of action were supported by either in vitro or in vivo models.

The present systematic review was conducted to explore the therapeutic potential of medicinal plants and herbal medicine in the treatment of NAFLD and steatohepatitis. The records for the review were collected from various scientific databases, such as Scopus, PubMed, Springer, NCBI, Google Scholar, Science Direct and Web of Science. The search was conducted using key words, such as ‘fatty liver’, ‘NAFLD’, ‘plants’, ‘medicinal’, ‘herbal medicine’ and ‘therapeutic uses’, and the data were limited to the period from 2016 to 2023. Inclusion criteria for the study were articles written in the English language, basic research studies and non-clinical studies. Conference abstracts, theses, case reports, reviews, commentaries and editorials were excluded. The author NEJ extracted the data, and both NEJ and SMJ independently screened all the retrieved abstracts using the inclusion and exclusion criteria. Any disagreements regarding inclusion were resolved through extensive discussion with the other two authors (RMS and CYC). Articles without liver histopathological assessment or positive drug and single-dose studies were excluded from the screening process.

After carefully applying the inclusion and exclusion criteria, a total of 55 articles were deemed relevant and selected for further analysis (as illustrated in Fig. 2). The information extracted from these articles was then organized and tabulated in an Excel spreadsheet. The key findings were subsequently summarized in three tables as follows: One detailing the traditional uses of the plants under investigation (Table II), the animal dietary model (Table III) and the other presenting the effects of these plants on NAFLD (Table IV).

It is noteworthy that >60% of the world's population, particularly in developing nations, relies mainly on medicinal plants for their healthcare needs. This renders traditional medicine a preferred healthcare system in a number of communities (26) due to its affordability, accessibility and low cost (27). As demonstrated in the present study, all the 20 herbs identified that were found to lead to a reduction in steatosis in liver histopathological analyses were traditionally used as herbal medicines for liver ailments, liver tonics, or for nourishing the liver (Table II) (28-62). This highlights the significance of traditional medicine in promoting liver health.

NAFLD is a condition characterized by the accumulation of excessive fat in the liver. This disease progresses from a simple state of liver steatosis to NASH and ultimately, into liver fibrosis, cirrhosis, and in severe cases, hepatocellular carcinoma (HCC). A review of the pathogenesis and histopathology of the disease in animals revealed that mice and rats are the commonly used models for the study of NAFLD (Table III). The C57BL/6 strain in mice, and the Wistar and Sprague-Dawley (SD) strains in rats are frequently used due to their inherent propensity to develop obesity, type 2 diabetes and NAFLD (63,64). The different stages of fatty liver are induced by altering the diet and chemicals used, including steatosis [confirmed by increased liver triglyceride levels, hepatocyte ballooning and Mallory bodies, also known as Mallory-Denk bodies (MDBs)], NASH, fibrosis and HCC, which are all dependent on the induction period.

When hepatocellular steatosis occurs with concurrent necro inflammatory reactions of the liver and hepatocellular ballooning with or without fibrosis and/or cirrhosis, it is diagnosed as NASH. Lobular inflammation and portal inflammation are both present in NASH, along with other histological lesions, such as hepatocellular ballooning, fibrosis, apoptotic bodies, sinusoidal collagen formation, MDBs, megamitochondria, glycogenated nuclei and iron deposition (64-66). The time of onset, as well as the degree of both NAFLD and accompanying metabolic features, are dependent on species, strain, sex, composition of the gut microbiota and the employed dietary intervention (67,68). Therefore, liver histology from animal models is crucial for elucidating the mechanisms and pathways involved in the pathogenesis of the NAFLD spectrum during the non-clinical stage. In clinical studies, regulatory agencies in the USA require liver histological endpoints in phase 3 studies (69). The ethnopharmacological usage of herbs may provide a reference with which to identify the appropriate animal dietary model. The simple steatosis or NASH observation from a high-fat diet (HFD) or methionine and choline deficiency (MCD) model may be suitable for this purpose.

HFD. The HFD model is the most frequently used dietary model for research in NASH. Research has shown that rats fed a HFD (70) containing 45-75 kcal% develop NASH after 12 weeks. These rats exhibit a phenotype similar to that of humans, characterized by obesity after 10 weeks, insulin resistance indicated by hyperinsulinemia and hyperlipidemia after 10 weeks, and glucose intolerance after 12 weeks (Table III). It is noteworthy that minimal fibrosis is only observable after 36-50 weeks of HFD (70).

MCD diet. The MCD diet is characterized by a high sucrose content and moderate fat content. This means that it typically contains 40% sucrose and 10% fat. However, this diet is deficient in two essential nutrients, choline and methionine. As a result, the ability of the body to oxidize fats and produce very low-density lipoprotein particles is impaired (71). This leads to the accumulation of fat in the liver, which can cause oxidative stress, liver cell death, inflammation and fibrosis after 8-10 weeks.

Notably, mice fed a MCD diet do not exhibit obesity, peripheral insulin resistance, or dyslipidemia (Table III), unlike humans with NASH. Instead, they experience significant weight loss, cachexia and low levels of serum insulin, fasting glucose, leptin, and triglycerides (72). The NASH phenotype with lobular inflammation and metabolic features, as well as ballooning, develops rapidly in these mice within 2-8 weeks (72).

Therefore, this model is suitable for studying NASH and its pharmacological treatment, but inadequate for studying NAFLD due to its multisystemic nature (73). Mouse strains exhibit varying responsiveness to an MCD diet (73).

Choline deficient L-amino acid-defined HFD. A HFD that is deficient in choline and amino acids can lead to the development of NASH with fibrosis in merely 6-9 weeks, even in the absence of significant weight loss (Table III). However, this diet does not fully replicate the metabolic syndrome observed in humans (74).

Streptozotocin (STZ) + HFD. When administered to mice, STZ has been found to damage the pancreatic islets and decrease insulin production. Additionally, a HFD diet beginning at 4 weeks of age, combined with the administration of neonatal STZ, has been shown to cause simple steatosis at 6 weeks (Table III), NASH at 8 weeks, and progressive pericellular fibrosis starting at 8-12 weeks, leading to HCC after 20 weeks (75).

Carbon tetrachloride (CCl₄) + HFD. Exposure to CCl₄ can trigger a response in the liver that leads to an accumulation of harmful lipid and protein peroxidation products, which can in turn, cause necrosis. When combined with a HFD, CCl₄ can exacerbate the development of NASH and fibrosis. In a previous study conducted with mice, it was found that multiple peritoneal injections of CCl₄ over a period of 4 weeks induced not only steatosis, but also hepatocellular ballooning, centrilobular fibrosis and hypertriglyceridemia; weight loss was also observed in the mice (76). Furthermore, the histological features worsened progressively with each administration of CCl₄ (76).

HFD, high-sugar diet and high-fat, high-fructose diet. Consuming fructose can significantly affect glucose and lipid metabolism, leading to several health issues, such as obesity, insulin resistance and lipid accumulation in the liver. Research conducted on rats and mice has demonstrated that drinking fructose-supplemented water for 8 weeks results in simple steatosis without NASH and contributes to obesity and insulin resistance (IR) (77). Additionally, rats that were administered a high-fat, high-fructose diet experienced hepatic inflammation after 16 weeks. Similar results were observed in rats that were fed a high-fat, high-sucrose diet (77). Of note, rats that were fed with glucose and sucrose exhibited a greater weight gain, but lesser hepatic fat accumulation as compared to a high fructose-fed diet (78).

Since NAFLD is associated with IR or obesity, the majority of the pathophysiological observations of the effect of these herbs listed in Table IV were shown to improve: i) lipid metabolism; ii) insulin resistance; iii) inflammation; iv) oxidative stress; and v) endoplasmic reticulum stress, in addition to the reduction of steatosis in the liver histological assessment. However, all rats with steatosis had elevated levels of liver injury markers, such as ALT and AST.

The research results from the 20 plants (Table IV) (28-121) related to markers for liver lipid metabolism revealed marked decreases in triglyceride levels ranging from 11.6 to 72.4%, total cholesterol levels from 13.1 to 50%, low-density lipoprotein levels ranging from 10.8-60.9%, and increases in high-density lipoprotein levels ranging from 13.2-58.2%. These plant extracts exert anti-hyperlipidemic effects that have the potential to reverse or reduce liver steatosis. In hyperlipidemia, the accumulation of high levels of cholesterol and triglycerides in the blood causes fat to accumulate in the liver, resulting in inflammation and oxidative stress (79). The highest dose evaluated and that was found to be effective in reducing liver fat ranged from 100-400 mg/kg body weight in rats (Table IV). Some of these plant extracts, such as Glossogyne tenuifolia (80) and Picrorhiza kurroa (53) leaves, have been found to be effective at a dose of 150 and 300 mg/kg, respectively. Moreover, neither of these extracts exhibited any signs of toxicity at the highest dose of 5 and 2 g/kg, respectively (Table IV). The extracts of Antidesma bunius (82,83), Aralia elata (84-86), Citrus aurantium (90-93), Curcuma longa Linn. (94-96), Cyclosorus terminans (101,102), Panax notoginseng (109), Pluchea indica (114), and Rosmarinus officinalis Linn (115,116) have also been shown to reduce the accumulation of lipids in rats fed a HFD through sterol regulatory element-binding transcription factor 1c (SREBP-1c), peroxisome proliferator-activated receptor-α (PPAR-α), fatty acid synthase (FAS), acetyl-CoA carboxylase (ACC), and carnitine palmitoyltransferase (CPT)2 regulation.

The adenosine monophosphate-activated protein kinase (AMPK) pathway plays a crucial role in regulating hepatic lipogenesis and β-oxidation and serves as a vital energy sensor for intracellular energy metabolism. It helps in regulating free fatty acids, de novo lipogenesis and hepatic lipid accumulation. The extracts of C. longa Linn. (94,95) and R. officinalis Linn. (115) can activate AMPK in NASH rats, leading to reduced dyslipidemia and hyperglycemia. Moreover, triggering AMPK signaling, affects adipocytokine production (122), which further highlights the potential of these extracts in regulating metabolic disorders.

Several plant extracts, including Aralia elata (84-86), Curcuma longa Linn. (94,95), Cyclosorus terminans (100-102), Panax notoginseng (109), and Pluchea indica (114) (Table IV), enhanced fatty acid β-oxidation by activating lipid antioxidant enzymes, such as CPT1 and peroxidation reduction through the activation of AMPK and PPAR-α.

Several plant extracts, including those derived from Hibiscus sabdariffa, Panax notoginseng, Antidesma bunius, Curcuma longa Linn., Aralia elata, Cyclosorus terminans, Pluchea indica, and R. officinalis Linn. (Table IV), upregulated the expression leels of SREBP-1c, FAS, ACC, and CPT-1, as well as PPAR-α, a regulator of β-oxidation, in rats that were fed a HFD (83,84,94,95,101,105,109,114,115). This suggests that AMPK activation by these herbal compounds is associated with de novo lipid synthesis, which is linked to the suppression of SREBP-1c, PPAR-α, FAS, ACC and CPT-2 expression. Moreover, these plant extracts promote the defense mechanism of β-oxidation, which leads to hepatic fatty acid depletion, by modulating CPT-1 and PPAR-α production.

Carbohydrate metabolism plays a crucial role in IR, as it is regulated by insulin hormone. When the insulin signal fails to prompt glucose absorption in cells due to IR, it leads to hyperglycemia, where glucose levels in the blood remain elevated. The body compensates for this by producing more insulin, which can cause hormonal imbalances and cell damage, particularly in liver cells. Excess insulin production can also lead to liver fat accumulation, which is a common occurrence in IR. Moreover, hyperglycemia and hormonal imbalances can cause inflammation and oxidative stress on liver cells, thus aggravating liver fat accumulation. Therefore, it is essential to manage IR through a healthy diet, regular exercise, and avoiding excessive alcohol consumption to prevent and treat NAFLD. IR leads to the accumulation of free fatty acids in the liver, which triggers de novo lipogenesis and causes NAFLD (123,124). Thus, the following studies have shown that certain natural extracts can help manage IR and its associated complications.

It is noteworthy that all the studies cited in Table IV (28-121), examining the effects of herbal plant extracts on IR, found significant decreases in glucose and insulin levels, as well as IR with the oral administration of these herbal extracts, namely, Abroma augusta (28,81). Aralia elata (84-86), Crocus sativus (97-99), Cyclosorus terminans (100-102), Glossogyne tenuifolia (80,103), Morus latifolia (107,108), Panax notoginseng (109), Pluchea indica (114), and Rubus ideaus (117,118), Trigonella foenum-graecum (119-121).

For instance, 200 mg/kg hexane extract of Cyclosorus terminans administered orally in rats fed a HFD was shown to reduce blood glucose levels, insulin and the homeostatic model assessment of insulin resistance (HOMA-IR) over a period of 2 weeks. This extract increased the expression of genes solute carrier family 2 member 4 (Slc2a4) and solute carrier family 2 member 2 (Scl2a2), thereby stimulating IR in the liver cells and soleus muscle. It also promoted the expression of insulin receptor substrate-1 (IRS1) and insulin receptor substrate-2(IRS2) genes, promoting hepatic and soleus muscle glycogen production (100).

Similarly, treatment with 1 g/kg aqueous Trigonella foenum-graecum bark extract for 28 days in rats reversed the effects of IR and lowered the HOMA-IR and apolipoprotein B (apoB) levels in the blood (119). IR is associated with increased secretion and decreased clearance of ApoB, which reduces low-density lipoprotein clearance (125). Trigonella foenum-graecum managed to reduce the effects of IR and ApoB, and increase LDL clearance (119).

The oral administration of extracts from various herbs, such as Antidesma bunius (83), Cassia obtusifolia (87), Crocus sativus (97), Cyclosorus terminans (101), Glossogyne tenuifolia (103), Hibiscus sabdariffa (105), Panax notoginseng (109) and Rubus ideaus (117), were found to lead to an improvement in the levels of inflammatory markers linked to liver damage. These herbal extracts have been found to reduce the levels of pro-inflammatory cytokines and inflammation. The effective dosages of these extracts range from 100-500 mg/kg body weight in rats, and no signs of toxicity were observed even at the highest evaluated dose of 2,000 mg/kg.

According to the study by Park et al (126), hepatic adiponectin induction reduced NASH-associated necro-inflammation and fibrosis by antagonizing TNF and regulating each other's secretion. Another study demonstrated that the oral administration of 500 mg/kg Hibiscus sabdariffa for 8 weeks reduced the release of pro-inflammatory cytokines, such as IL-6 and TNF-α, which inhibited the development of NASH (105).

Liver inflammation causes inflammatory damage by increasing lipid accumulation and redistribution from adipose tissue to the liver. Hepatic steatosis, steatohepatitis, and fibrosis are the first steps in the evolution of NAFLD caused by liver inflammation (127). Herbal extracts have been found to protect against the advancement of hepatic steatosis to steatohepatitis by reducing liver inflammation. This is achieved by the suppression of inflammatory signaling pathways, controlling dyslipidemia, and enhancing liver function in patients with NAFLD (127). Herbal remedies, such as Antidesma bunius (83), Cassia obtusifolia (87), Crocus sativus (97), Cyclosorus terminans (101), Glossogyne tenuifolia (103), Hibiscus sabdariffa (105), Panax notoginseng (109) and Rubus ideaus L. (117), and (Table IV) have been shown to reduce the expression levels of hepatic inflammatory cytokines (TNF-α, IL-6, IL-8 and IL-1β) and to further ameliorate liver fibrosis. Hou et al (109) demonstrated that 30 mg/kg Panax notoginseng ethanol extract reduced the levels of inflammatory markers, namely TNF-α, IL-6, IL-8, IL-1 and IL-1β, and no signs of toxicity were observed at the highest chronic dose evaluated at 1,200 mg/kg for 28 days. From the results presented in Table IV, some of these herbs reduced the biomarker for insulin resistance and lipid metabolism in addition to steatosis reduction. As insulin resistance and lipid accumulation induced steatosis, most likely these herbs will ameliorate IR and lipid accumulation before the development of steatosis. An earlier study on Hibiscus sabdariffa aqueous extract at a lower dose of 300 mg/kg for 10 weeks, demonstrated reduced weight gain or obesity in rats fed a HFD through the inhibition of adipogenesis (128), indicating a close link to metabolic syndrome.

A total of eight herbal plants, namely Abroma augusta (28,81), Antidesma bunius (82,83), Cassia obtusifolia (87-89), Curuma longa (94-96), Crocus sativus (97-99), Hibiscus sabdariffa (104,105), Rubus ideaus (117,118), and Trigonella foenum-graecum (119-121)_ were found to be safe at the highest toxicity dose evaluated and significantly reduce the levels of pro-oxidants, namely, malondialdehyde, ER stress, ROS and oxidative end products, such as 4-hydroxynonenal (Table IV).

Visceral fat accumulation can lead to oxidative stress, which is a common characteristic of NAFLD. This, in turn, triggers lipid peroxidation, causing oxidative damage throughout the body (105). The development of NAFLD can result in liver damage due to an imbalance between the production of reactive species and antioxidant defense. NAFLD affects lipid metabolism, leading to the production of ROS through fatty acid oxidation. The effective extract dosages ranged from 200-1,500 mg/kg body weight in experimental rats, and all of these plant extracts did not lead to any signs of toxicity at the highest evaluated dose of 2,000 mg/kg (Table IV). The ethanol extract of 200 mg/kg Rubus ideaus has been shown to significantly decrease the level of malondialdehyde, while increasing the levels of the antioxidants, superoxide dismutase, glutathione and glutathione peroxidase, thereby reducing liver oxidative stress (117). The extracts of Aralia elata (84-86), Curcuma longa (94-96), Cyclosorus terminans (100-102), Panax notoginseng (109), and Pluchea indica (114) have been found to significantly promote fatty acid oxidation through the activation of PPAR-α and PPAR-γ.

The pathological disorders associated with obesity and NAFLD include the ER stress response as one of the primary characteristics (42). One of the primary characteristics of these disorders is the activation of the ER stress axis (42). However, Curcuma longa Linn. has been found to block this axis by activating AMPK, a physiological regulator of the mTOR signaling pathway that helps lower lipid metabolism. By activating AMPK, Curcuma longa Linn. can reduce the ER stress response (94). This natural herb has also been found to prevent hepatic dyslipidemia by downregulating the levels of phosphorylated mammalian target of rapamycin (p-mTOR), phosphorylated ribosomal protein s6 kinase (p-S6K) and phosphorylated eukaryotic translation initiation factor 4E-binding protein 1 (p-4-EBP-1), while alleviating ER stress (95). Curcuma longa Linn. has been proven to inhibit overnutrition-induced hepatic lipid accumulation, by controlling SREBP-1 and FAS through the protein endoplasmic reticulum kinase/eukaryotic translation initiation factor 2, which regulates lipid metabolism (95). Additionally, Curcuma longa Linn. has been found to modulate ER stress response and redox imbalance by impacting mTORC1, demonstrating the role of the mechanistic target of rapamycin complex 1 (mTORC1) activation and protein folding in the pathogenic process of hepatic dyslipidemia (94).

The limitations of the studies identified in the present systematic review were the usage of non-standardized extracts, which could affect the reproducibility of their therapeutic effects. Variation in the bioactive synthesis is expected with herbs collected or planted at different locations as the biosynthesis of these bioactive compounds is dependent on the quality of the soil, altitude, environment and the time of harvest.

In conclusion, in the present systematic review, 20 herbs were found to reduce steatosis in histopathological assessment. Of these, 15 herbs, namely A. augusta (28,81), A. bunius (82,83), A. elata (84-86), C. obtusifolia (87-89), C. aurantium (90-93), C. longa (94-96), C. sativus (97-99), C. terminans (100-102), G. tenuifolia (80,103), H. sabdariffa (104,105), P. notoginseng (109), P. indica (114), R. officinalis (115,116), R. ideaus (117,118) and T. foenum-graecum (119-121) exhibited a mode of action involving lipid metabolism, insulin resistance and/or inflammatory markers. The remaining five herbs, namely M. oleifera (46,106), M. latifolia (107,108), P. emblica (110,111), P. kurroa (53,112) and P. anisum (113) warrant further investigations to establish their cross-link mode of action. These herbs exhibited no signs of toxicity. The herbs were found to exert a positive effect against NAFLD, and understanding its mechanism of action is probably useful for the treatment of other metabolic diseases associated with NAFLD, namely dyslipidemia, diabetes, or hypertension. Further studies are required to identify the bioactive compounds and standardize the extracts. Planning for either in vivo or in vitro experimental studies is essential to identify the bioactive compounds present in herbs. The traditional usage of these herbs can provide a useful reference for such studies. This process helps in understanding the bioactive components of the herbs and standardizing the extracts to ensure reproducible therapeutic effects. An understanding of the mechanism of action of these herbs is useful for planning more successful clinical trials. The development of NAFLD is time-consuming and herbs may provide alternative treatment to its prevention.