The development and progression of MAFLD are closely linked to gut dysbiosis. Alterations in the gut microbiota exhibit taxonomic-level specificity and dynamic changes across disease stages ( Table 1 ).

A common hallmark of dysbiosis is an increased abundance of Proteobacteria and Actinobacteria phyla, along with an elevated Firmicutes/Bacteroidetes ratio (Loomba et al., 2019). At the phylum level, MAFLD patients typically show reduced abundance of Bacteroidetes and increased abundance of Firmicutes and Proteobacteria (Forlano et al., 2022), forming a characteristic “Firmicutes/Bacteroidetes imbalance.” In healthy individuals, Firmicutes and Bacteroidetes dominate, while Proteobacteria and others are relatively scarce (Jennison and Byrne, 2021). Disruption of this homeostasis may drive MAFLD progression by promoting energy harvest and inflammatory responses. At the family level, MAFLD patients exhibit an increased relative abundance of Enterobacteriaceae and a decrease in Rikenellaceae and Ruminococcaceae, which possess anti-inflammatory metabolic functions (Chen and Vitetta, 2020). Notably, Ruminococcaceae abundance shows a positive association with significant fibrosis, suggesting its dynamic changes correlate with disease severity (Boursier et al., 2016). However, investigations into Ruminococcaceae abundance in MAFLD/MASH patients reveal inconsistent trends across populations. Del Chierico et al. (2017) observed a significant increase in Ruminococcaceae in children/adolescents with MAFLD or MASH compared to healthy controls. Conversely, a meta-analysis by Li et al. (2021) involving 1,265 subjects (including 577 MAFLD patients from 8 countries) found reduced Ruminococcaceae in MAFLD patients. Lee et al. (2020) further highlighted this discrepancy, reporting a negative association between Ruminococcaceae abundance and significant fibrosis in non-obese patients—a finding contradictory to Boursier et al (Boursier et al., 2016). These collective data indicate that Ruminococcaceae abundance varies significantly depending on patient cohorts and metabolic subgroups. Furthermore, alterations at the genus level are more complex: pro-inflammatory genera such as Escherichia and Dorea expand, while butyrate-producing genera like Faecalibacterium, Coprococcus, and Prevotella are significantly reduced (Aron-Wisnewsky et al., 2020). The abundance change of Eubacterium rectale is particularly unique—it increases in moderate-to-severe MAFLD but decreases sharply when fibrosis develops, suggesting a dual role in compensatory adaptation and profibrotic processes across different pathological stages (Chen and Vitetta, 2020).

As the disease progresses to metabolic dysfunction-associated steatohepatitis (MASH) and fibrosis, functional remodeling of the microbiota intensifies. Metagenomic analysis reveals abnormal activation of tryptophan/phenylalanine metabolism and lipopolysaccharide (LPS) biosynthesis pathways in MASH-associated microbiota, while pathways for cellulose degradation and short-chain fatty acid (SCFA) synthesis (e.g., butyrate) are suppressed (Boursier et al., 2016). This metabolic shift amplifies endotoxemia and oxidative stress via the gut-liver axis, further worsening insulin resistance. Non-invasive diagnostic techniques based on microbial signatures are rapidly advancing; for instance, a 16-genus marker model including Ruminococcus and Streptococcus significantly improves diagnostic accuracy for advanced fibrosis (Loomba et al., 2019; Oh et al., 2020). Methodologically, targeted 16S rRNA sequencing is commonly used for bacterial community analysis, while 18S rRNA or internal transcribed spacer (ITS) sequencing can profile fungal communities; metagenomic sequencing (mNGS) and probe-capture techniques enhance the detection of low-abundance species. Studies indicate that gut microbial α-diversity in MAFLD patients decreases with increasing hepatic fat accumulation, and meta-analyses reveal a core dysbiotic signature characterized by increased Escherichia and Prevotella, alongside decreased Faecalibacterium and Ruminococcaceae (Long et al., 2024). These findings suggest that hierarchical disruptions in microbial composition and function are not only biomarkers for MAFLD but also key pathological drivers of disease progression.

MAFLD is characterized by excessive hepatic triglyceride accumulation (Rong et al., 2023b). Its pathological progression is closely linked to gut-liver axis dysfunction driven by gut dysbiosis. The gut-liver axis forms a bidirectional regulatory network via the portal circulation, bile acid metabolism, and immune signaling (Albillos et al., 2020; Wu et al., 2022; Xiang et al., 2023; Siddiqui et al., 2025). Dysregulation of microbial metabolites, gut barrier impairment, and bile acid signaling imbalance constitute three core mechanisms driving hepatic lipid metabolism abnormalities, inflammation activation, and fibrosis (Blesl and Stadlbauer, 2021) ( Figure 2 ).

Recent studies have revealed the multifaceted mechanisms by which gut microbiota and their metabolic regulation contribute to treating MAFLD. Lactobacillus and Bifidobacterium significantly reduce serum cholesterol levels by modulating host metabolic pathways, likely through inhibiting intestinal cholesterol absorption and promoting bile acid excretion (Wu and Chiou, 2021). Further research indicates functional differentiation in liver farnesoid X receptor (FXR) subtypes during lipid metabolism regulation. FXRα2 exhibits stronger triglyceride (TG)-inhibiting capacity than FXRα1 via specific binding to DNA motifs, suggesting that targeted selective activation of FXR subtypes may become a novel therapeutic strategy for MAFLD (Ramos Pittol et al., 2020).

In probiotic combination interventions, a mixture of six probiotics (including Lactobacillus and Bifidobacterium) significantly increased the abundance of beneficial bacteria such as Agathobaculum, Blautia, and Ruminococcus in the gut while reducing hepatic free fatty acids (FFA) and body mass index (BMI), demonstrating the synergistic role of microbiota in ameliorating metabolic disorders (Fang et al., 2022). Exercise intervention reshapes gut microbiota structure, such as reducing Parabacteroides and Flavobacterium, to enhance hepatic fatty acid oxidation capacity. Independent of weight loss, exercise suppresses the NF-κB inflammatory pathway, thereby reducing intrahepatic lipid accumulation (Ortiz-Alvarez et al., 2020).

For targeted microbial therapies, specific probiotic strains like Lactobacillus rhamnosus GG (LGG) inhibit intestinal NF-κB signaling to reduce systemic inflammation, while their metabolites activate the FGF21-adiponectin axis to promote lipid metabolism (Liu et al., 2020) and stimulate butyrate-producing bacteria proliferation to repair the gut barrier (Zhao et al., 2019). Lactococcus lactis subsp. cremoris outperforms LGG in ameliorating high-fat-induced metabolic dysregulation, evidenced by reduced serum cholesterol, attenuated hepatic steatosis, and restored glucose tolerance (Naudin et al., 2020). The multi-strain probiotic VSL#3 alleviates liver inflammation by suppressing the NF-κB pathway and downregulating key lipogenesis genes (SREBP-1c and FAS) (Jena et al., 2020). Prebiotics and synbiotics not only enhance fatty acid β-oxidation by upregulating PPAR-α/CPT-1 but also inhibit colonization of pro-inflammatory bacteria such as Enterobacteriaceae, thereby improving insulin resistance and liver injury (Alves et al., 2017). These findings highlight the potential of precision intervention strategies based on microbiota-host interactions in MAFLD management.

Current clinical research on fecal microbiota transplantation (FMT) for MAFLD remains exploratory. Three key trials reveal its potential and limitations: Craven et al. (2020) found that allogeneic FMT significantly improved intestinal permeability in MAFLD patients, though without improving HOMA-IR or MRI-PDFF. Witjes et al. (2020) demonstrated that FMT from healthy donors upregulated hepatic ARHGAP18 (a cytoskeleton regulator) and serine dehydratase (SDS) expression in patients with MASH while reducing serum GGT and ALT. Xue et al. (2022) reported decreased CAP values post-FMT alongside proliferation of butyrate-producing bacteria, activation of the FXR/TGR5 axis, and inhibition of fatty acid synthase (FASN). These results suggest FMT may mitigate liver injury by repairing the gut barrier, regulating host gene expression, and modulating metabolic pathways. However, heterogeneous efficacy, long-term safety concerns, and insufficient mechanistic validation (Qiu et al., 2024) require resolution through standardized donor screening and optimized trial designs.

Emerging gut-liver axis strategies indicate that symbiotic supplementation enriches butyrate-producing microbiota, elevates short-chain fatty acid (SCFA) levels, improves insulin resistance, inhibits hepatic lipogenic enzymes, and alleviates inflammation/oxidative stress via FXR/TGR5 signaling (Eslamparast et al., 2014). This underscores the potential of microbiota modulation to reshape gut-liver metabolic crosstalk, offering a microbe-centric paradigm for MAFLD.

The field of microbiota-targeted therapy is evolving from single-strain supplementation toward systematic ecological modulation. Future advances demand prioritizing functional gene clusters over individual species, establishing real-time monitoring of dynamic microbiota-host interactions, and leveraging synthetic biology to design therapeutic artificial microbial ecosystems. We prioritize butyrate synthesis (e.g., but/buk gene clusters) (Kalkan et al., 2025)and bile acid metabolism (e.g., bai/bsh genes) (Li et al., 2023) as core therapeutic targets due to their direct regulation of intestinal barrier integrity, host immunity, and metabolic homeostasis; concurrently, short-chain fatty acid transporters and antimicrobial peptide synthesis gene clusters will be incorporated to enhance microbial colonization resistance. Clinical efficacy will be evaluated via a multidimensional strategy: metagenomic tracking of functional gene abundance, metabolomic quantification (GC-MS/LC-MS) of butyrate and bile acid metabolites, host-response analysis of serum inflammatory markers and intestinal barrier indicators, and systematic correlation with clinical symptom scores to validate therapeutic mechanisms and translational potential. Ultimately, by redefining the microbiome as a programmable biological network, precise strategies for MAFLD prevention and treatment can be achieved.

Metabolomic studies reveal significant metabolic dysregulations in patients with MAFLD, involving multiple pathways such as lipid, amino acid, bile acid, and energy metabolism. These alterations are closely linked to disease progression. MAFLD patients commonly exhibit hepatic lipid deposition, characterized by elevated free fatty acid (FFA) levels (Guo et al., 2022), increased triglyceride/high-density lipoprotein cholesterol (TG/HDL-C) ratio (a non-invasive diagnostic marker) (Fan et al., 2019), and phospholipid imbalance (e.g., decreased phosphatidylcholine/phosphatidylethanolamine (PC/PE) ratio) (Peng et al., 2021). Impaired hepatic mitochondrial β-oxidation leads to long-chain fatty acid accumulation, exacerbating lipotoxicity (Koliaki et al., 2015).

Dysregulated branched-chain amino acid (BCAA) metabolism is a hallmark feature, with elevated blood levels of BCAAs (e.g., leucine, isoleucine) and their metabolites correlating with insulin resistance and hepatic steatosis (Lo et al., 2022). Concurrently, increased aromatic amino acids (e.g., phenylalanine, tyrosine) and glutamate may promote inflammation and fibrosis via mTOR pathway activation (Samuel et al., 2004). Gut microbiota dysbiosis (e.g., elevated Firmicutes/Bacteroidetes ratio) (Jasirwan et al., 2021) (Boursier et al., 2016) disrupts the gut-liver axis through bile acid metabolism, resulting in increased secondary bile acids (e.g., deoxycholic acid) and reduced primary bile acids (e.g., taurocholic acid). This impairs farnesoid X receptor (FXR) signaling, worsening lipid dysregulation and inflammation (Wang et al., 2025).

While mitochondrial adaptive responses persist in simple steatosis (e.g., compensatory enhanced fatty acid oxidation), progression to MASH reduces oxidative phosphorylation efficiency. Accumulation of tricarboxylic acid (TCA) cycle intermediates (e.g., citrate, succinate) and elevated reactive oxygen species (ROS) production drive cellular damage and fibrosis (Koliaki et al., 2015).

Metabolomics has identified multiple potential biomarkers ( Table 2 ), including serum BCAAs, 2-aminoadipic acid (2-AAA), and specific lipid profiles (e.g., Lys phosphatidylcholines), which correlate significantly with hepatic fat content, inflammation, and fibrosis severity (Di Mauro et al., 2021). Integrating these with machine learning models (e.g., laboratory parameter-based MAFLD screening) (Yip et al., 2017) or traditional scoring systems (e.g., BARD score) (Rigor et al., 2022) enhances diagnostic and staging accuracy.

Thus, metabolomics not only provides molecular insights into MAFLD pathogenesis but also enables novel approaches for non-invasive diagnosis, disease subtyping, and targeted therapies (e.g., FXR agonists, gut microbiota modulation) (Xu et al., 2025). Future research should integrate multi-omics data to precisely delineate metabolic network dynamics in MAFLD progression.

Metabolomics systematically analyzes dynamic changes in endogenous metabolites to elucidate the pathological mechanisms of MAFLD. This metabolic disorder, characterized by hepatic lipid accumulation, involves complex pathogenesis encompassing dysregulated lipid, amino acid, and carbohydrate metabolism. Metabolomics thus provides novel insights into these disturbances ( Figure 3 ).

The core etiology of MAFLD stems from disrupted hepatic lipid metabolism, primarily characterized by excessive triglyceride (TG) accumulation. This steatosis develops when lipid metabolic capacity becomes overwhelmed due to an imbalance between lipogenesis and degradation pathways, resulting in abnormal lipid deposition within hepatocytes (Chen et al., 2019). Excessive hepatic free fatty acid (FFA) accumulation serves as a key driver of this process, originating from three interconnected sources: adipose tissue lipolysis increasing circulating FFA levels (closely linked to insulin resistance [IR] and hepatic inflammation) (Griffin et al., 1999); hyperactive de novo lipogenesis (DNL) where skeletal muscle IR-induced hyperglycemia and hyperinsulinemia activate transcription factors ChREBP and SREBP1c, upregulating lipogenic enzymes that convert excess glucose into FFA (Samuel and Shulman, 2018); and dietary lipids entering the liver through bile acid receptor-mediated absorption (e.g., via FXR), further exacerbating FFA burden (Cheng et al., 2024).FFA overaccumulation not only impairs mitochondrial β-oxidation (e.g., through CPT1 downregulation (Serviddio et al., 2011)) but also promotes oxidative stress and hepatic fibrosis (Ramanathan et al., 2022). Beyond FFA dysregulation, other lipid abnormalities contribute to MAFLD progression. An imbalanced phosphatidylcholine-to-phosphatidylethanolamine (PC/PE) ratio disrupts membrane integrity, with decreased ratios distinguishing simple steatosis from MASH and liver injury (Peng et al., 2021). Mitochondrial adaptive responses (e.g., PPARα and CPT1 upregulation) may initially enhance fatty acid oxidation, but these compensatory mechanisms progressively fail amid evolving IR and hormonal changes like leptin dysregulation (Begriche et al., 2013).

Amino acid metabolic disturbances critically influence MAFLD pathogenesis, particularly the branched-chain amino acid (BCAA) and aromatic amino acid (AAA) imbalance. BCAA dysregulation activates the mammalian target of rapamycin (mTOR) pathway, exacerbating IR and hepatocyte steatosis (Lo et al., 2022). BCAT2 knockout models demonstrate that BCAA accumulation induces energy metabolism disorders while paradoxically conferring obesity resistance, revealing its dual metabolic roles (Ananieva et al., 2017). Clinical studies confirm significant positive correlations between serum BCAA levels and intrahepatic triglyceride content (IHTC), ALT, AST, and GGT in MAFLD patients. Critically, the BCAA-IHTC association persists after adjusting for obesity and IR, indicating BCAA’s direct steatogenic role (Ni et al., 2023) (van den Berg et al., 2019). In vitro validation shows valine upregulates lipogenic transcription factors (e.g., SREBP-1c), promoting lipid synthesis while inhibiting fatty acid oxidation to increase hepatocellular TG (Ni et al., 2023). Notably, obese MAFLD patients exhibit higher BCAA elevations than non-obese counterparts, with valine and isoleucine accumulation directly correlating with hepatic fat content (Lischka et al., 2020). High BCAA intake also correlates with liver injury severity in obese MAFLD patients, highlighting diet-metabolism interactions (Galarregui et al., 2021).

AAA metabolic abnormalities associate closely with hepatic inflammation and fibrosis in MASH, potentially through pro-inflammatory pathway activation (e.g., NF-κB) (Kalhan et al., 2011). Excessive glutamine breakdown causes ammonia accumulation, impairing hepatocyte function. Recent evidence reveals ammonia promotes SREBP-1 maturation and lipogenesis by activating SCAP/Insig complex dissociation, elucidating its molecular role in MAFLD/MASH (Cheng et al., 2022). This process interfaces with gut microbiota metabolism, as elevated serum BCAA correlates with dysbiosis and IR (Pedersen et al., 2016), positioning the “gut microbiota-amino acid-liver” axis as central to MAFLD. Collectively, amino acid dysregulation orchestrates MAFLD pathology by modulating lipid synthesis, inflammation, and energy metabolism.

Carbohydrate metabolism dysregulation represents another hallmark of MAFLD, manifesting through coordinated glycolysis and gluconeogenesis dysfunction. Elevated blood lactate and pyruvate in MAFLD patients indicate disordered hepatic glucose metabolism and mitochondrial impairment (Koliaki et al., 2015). Dietary patterns critically drive this imbalance: high-glycemic-index (GI) diets induce postprandial hyperglycemia, stimulating hepatic DNL and lipid accumulation (Parker and Kim, 2019). Excessive monosaccharide/disaccharide intake (e.g., fructose, sucrose, glucose) promotes MAFLD progression primarily through ChREBP activation (Katz et al., 2021). As a central lipogenic transcription factor, ChREBP directly binds promoters of DNL enzymes like fatty acid synthase (FASN) and acetyl-CoA carboxylase (ACC) (Ortega-Prieto and Postic, 2019). High-carbohydrate diets enhance ChREBP nuclear translocation and synergism with SREBP-1c, driving postprandial metabolic reprogramming (Linden et al., 2018). In 30%-sucrose-fed mouse models, aberrant ChREBP activation increases hepatic lipid droplets and inflammatory signaling—phenotypes partially reversed by ChREBP inhibition (Daniel et al., 2021). ChREBP also mediates fructose-induced gluconeogenesis dysregulation via insulin-independent mechanisms, indicating its unique role in metabolic compensation (Kim et al., 2016).

Investigation of these pathological mechanisms reveals that MAFLD’s metabolic disturbances involve multidimensional crosstalk. Insulin resistance acts as the central hub, coordinating synergistic dysregulation across lipid, amino acid, and carbohydrate metabolism to promote concurrent hepatocellular injury, inflammation, and fibrogenesis—ultimately driving progression from steatosis to MASH.

Metabolomics research provides a systemic perspective for elucidating the pathogenesis of MAFLD and developing clinical interventions, driving a paradigm shift from single-pathway targeting toward systemic network modulation. At the foundational intervention level, scientific dietary management remains central: low-fat, high-fiber diets alleviate intrahepatic lipid deposition by optimizing metabolic profiles, while ω-3 polyunsaturated fatty acid (EPA/DHA)-rich regimens significantly reduce hepatic triglycerides, enhance insulin sensitivity, and suppress inflammation (Scorletti and Byrne, 2018). The Mediterranean diet, rich in olive oil, nuts, and deep-sea fish, demonstrates efficacy by modulating lipid metabolism, reducing liver enzymes such as ALT and AST, and attenuating hepatic inflammation (Gantenbein and Kanaka-Gantenbein, 2021). More recently, the ketogenic diet—characterized by very low carbohydrate and high fat intake—has been shown to improve MAFLD through enhanced lipid oxidation and reduced hepatic lipogenesis (Watanabe et al., 2020). Exercise functions as a synergistic metabolic modulator, improving glucose-lipid metabolism and reducing intrahepatic fat content (Vanweert et al., 2021), with combined resistance and aerobic training yielding superior outcomes in both non-obese and obese MAFLD patients (Zhang et al., 2022).

Pharmacological strategies for MAFLD and metabolic dysfunction-associated steatohepatitis (MASH) exhibit multi-tiered advances. Classic insulin sensitizers like metformin improve underlying metabolic abnormalities by regulating glucose-lipid metabolism, though evidence for histological improvement such as fibrosis reversal in MASH remains limited (Ruan et al., 2023). Conversely, the PPARγ agonist pioglitazone significantly reduces hepatic steatosis, lobular inflammation, and hepatocyte ballooning in non-diabetic MASH patients while delaying diabetes progression (Cusi et al., 2016). Among emerging targeted agents, the bile acid-fatty acid conjugate Aramchol inhibits SCD1 to reduce lipid synthesis, with its Phase III ARMOR trial (NCT04104321) for F2-F3 fibrosis MASH patients currently evaluating efficacy (Alkhouri et al., 2021). The FXR agonist Obet cholic acid (OCA), a selective bile acid modulator, significantly improved MASH-related fibrosis (≥1-stage improvement without worsening) at 25 mg/day in Phase III trials, though approximately 20% of patients discontinued treatment due to pruritus (Chiang and Ferrell, 2022). Notably, the GLP-1 receptor agonist semaglutide demonstrated substantial advantages in a Phase II trial where 0.4 mg daily treatment for 72 weeks achieved histological resolution without worsening fibrosis in 320 MASH patients, positioning it as the most promising metabolic-regulating therapy to date (Zhang et al., 2025). For severely obese patients, foregut bariatric surgery is recommended by international guidelines as an effective intervention (European Association for the Study of the Liver (EASL) et al., 2016), significantly improving BMI, fibrosis scores, and histological features (Nickel et al., 2018), while statins serve as adjunctive therapy for dyslipidemia comorbidities but remain contraindicated in decompensated cirrhosis (Chalasani et al., 2018).

In summary, the current therapeutic framework integrates foundational lifestyle interventions, precision medications targeting the gut-liver axis such as OCA and semaglutide, and surgical approaches, highlighting the necessity for metabolomics-driven individualized treatment selection. Semaglutide demonstrates superior histological resolution and safety profiles, whereas OCA improves fibrosis but faces limitations due to side effects. Future exploration of combination strategies—particularly GLP-1 and FXR agonist synergism—is warranted to cooperatively regulate multiple pathological pathways and optimize therapeutic outcomes.