Regular exercise significantly enriches Firmicutes phylum bacteria, particularly SCFAs-producing families like Ruminococcaceae, Lachnospiraceae, and Erysipelotrichaceae (74, 79, 80). Key functional examples include the enrichment of Faecalibacterium prausnitzii (Ruminococcaceae) in high-intensity female athletes and endurance runners, whose butyrate production promotes intestinal gluconeogenesis and glucose homeostasis (71). Similarly, Lachnospiraceae genera (Dorea, Coprococcus, Roseburia) demonstrate increased abundance in regular exercisers, producing acetate and butyrate to regulate blood glucose, mitigate allergic responses, and enhance gut immunity (81, 82). Marathon athletes further exhibit high enrichment of Veillonella, which utilizes lactate to produce SCFAs (83).

The resulting elevation in butyrate and propionate levels improves PCOS metabolic disturbances through three primary pathways. First, butyrate acts as a histone deacetylase (HDAC) inhibitor (84, 85), reducing the activity of HDAC1/2. This regulates chromatin accessibility and affects gene expression associated with processes such as androgen synthesis (e.g., CYP17A1 (32)) and gluconeogenesis. Second, propionate activates Free Fatty Acid Receptor 3 (FFAR3) on intestinal L-cells, stimulating GLP-1 secretion to enhance insulin sensitivity. GLP-1 enhances insulin secretion from pancreatic β-cells, inhibits glucagon release, delays gastric emptying, and promotes satiety, collectively contributing to improved glucose homeostasis and insulin sensitivity. Third, as the primary energy source for colonocytes, butyrate increases epithelial proliferation, upregulates tight junction proteins (ZO-1, Occludin), and reduces intestinal permeability (86, 87), preventing LPS translocation and subsequent systemic inflammation. Collectively, these SCFAs-mediated effects ameliorate metabolic dysfunction, inflammation, and barrier integrity, providing a multi-faceted foundation for PCOS symptom improvement.

The efficacy of exercise in modulating gut microbiota and metabolic health is indeed dose-dependent, with intensity serving as a critical modulator. While the term “high-intensity” applied to athletes denotes a level of training far exceeding general recommendations, the exercise intensity beneficial for PCOS management is both quantifiable and subject to an optimal range rather than a simple upper limit. Prescriptions are typically grounded in objective physiological metrics to ensure safety and efficacy. Common quantifiable measures include: (1)Percentage of Maximum Heart Rate (%HRmax): Moderate-intensity is defined as 64-76% HRmax; vigorous-intensity as 77-93% HRmax (88); (2)Percentage of Heart Rate Reserve (%HRR) or Oxygen Uptake Reserve (%VO₂R): Moderate-intensity corresponds to 40-59% HRR/VO₂R, while vigorous-intensity to 60-89% (89, 90); (3)Rating of Perceived Exertion (RPE): Using the Borg scale (6-20), moderate intensity targets 12-13 (“somewhat hard”), and vigorous intensity targets 14-16 (“hard” to “very hard”) (91).

For PCOS populations, evidence suggests that vigorous-intensity exercise (e.g., High-Intensity Interval Training(HIIT) protocols involving bouts at 80-90% HRmax) is highly effective for improving insulin sensitivity and cardiorespiratory fitness (76). However, the optimal intensity must be individualized. Factors such as baseline fitness, PCOS phenotype, joint health, and exercise tolerance necessitate personalization. The prevailing consensus recommends beginning with moderate-intensity exercise for sedentary individuals, progressively incorporating vigorous intervals as tolerance improves, while strictly avoiding excessive intensity that could provoke a negative stress response (92). This structured, quantifiable approach ensures that exercise intervention remains a potent, evidence-based strategy for gut microbiota and metabolic remodeling in PCOS.

Exercise regulates the gut-liver axis BA pool composition and function, increasing BA excretion (93, 94). By elevating the Firmicutes/Bacteroidetes ratio (55, 56, 67), exercise promotes microbial conversion of primary to secondary BAs. Exercise promotes this shift towards secondary BAs primarily by enriching gut bacteria harboring bile salt hydrolase (BSH) and 7α-dehydroxylase activities (e.g., certain Firmicutes like Clostridium cluster XIVa and XVI), thereby enhancing microbial conversion of primary to secondary BAs. These secondary BAs act as endogenous ligands for the FXR, recruiting nuclear receptor coactivators to activate FXR signaling with dual organ effects. The activation of FXR signaling in both liver and intestine represents a master regulator of systemic glucose and lipid metabolism, integrating microbial BA metabolism with host metabolic homeostasis. In the liver, FXR activation suppresses cholesterol 7α-hydroxylase (CYP7A1) expression to curb excessive BA synthesis while concurrently upregulating fibroblast growth factor 19 (FGF19) secretion, further inhibiting synthesis via negative feedback. Intestinally, FXR activation maintains glucose homeostasis by inhibiting hepatic gluconeogenesis enzyme expression (95) and enhances hepatic detoxification capacity. This axis provides a molecular framework for tissue-specific, multi-organ regulation of metabolic balance.

Exercise reduces Proteobacteria abundance (96–98), with regular moderate-intensity resistance training specifically lowering circulating LPS levels (99). Crucially, exercise-induced increases in SCFAs, particularly butyrate, play a fundamental role in fortifying the intestinal barrier. Butyrate serves as the primary energy source for colonocytes, stimulating epithelial proliferation and upregulating tight junction proteins (ZO-1, Occludin), thereby significantly reducing intestinal permeability and preventing LPS translocation. This reduction disrupts inflammatory cascades through coordinated mechanisms. Diminished LPS impairs Toll-like receptor 4 (TLR4)/Myeloid differentiation primary response 88 (MyD88) complex formation, blocking nuclear factor kappa B (NF-κB) phosphorylation and subsequent pro-inflammatory cytokine transcription. Concurrently, SCFAs activate G-protein coupled receptor 43 (GPR43), inhibiting apoptosis-associated speck-like protein (ASC) oligomerization, caspase-1 activation, and NLRP3 inflammasome assembly. These changes collectively promote macrophage polarization from pro-inflammatory M1 to anti-inflammatory M2 phenotypes.Furthermore, SCFAs, particularly butyrate, contribute to the activation of nuclear factor erythroid 2-related factor 2 (Nrf2) by inhibiting its cytoplasmic repressor Kelch-like ECH-associated protein 1 (KEAP1). This inhibition stabilizes Nrf2 and facilitates its translocation to the nucleus, where it binds to antioxidant response elements (ARE) in the promoter regions of target genes, boosting the transcription of antioxidant enzymes (e.g., heme oxygenase-1, NAD(P)H quinone dehydrogenase 1, glutathione peroxidase) and enhancing oxidative stress resistance (100, 101).

The three core mechanistic axes—SCFAs network reconstruction, BA-FXR signaling activation, and LPS-inflammation inhibition—do not operate in isolation but engage in extensive crosstalk, forming a dynamic and integrated metabolic network that underlies exercise’s ameliorative effects on PCOS.

SCFAs & BA/FXR: SCFAs, particularly butyrate, acting as HDAC inhibitors, may influence the epigenetic regulation of genes involved in BA metabolism (e.g., CYP7A1, FXR) and FXR target genes, potentially amplifying FXR signaling efficacy. Conversely, FXR activation in the intestine can modulate the expression of genes involved in mucosal defense and potentially influence the niche for SCFAs-producing bacteria.

SCFAs & LPS/Inflammation: Butyrate’s enhancement of gut barrier integrity is fundamental to reducing LPS translocation and systemic inflammation. Simultaneously, reduced inflammation (via LPS inhibition and SCFAs-GPR43 signaling) creates a more favorable gut environment for beneficial SCFAs-producing bacteria to thrive. SCFAs also directly suppress NLRP3 inflammasome activation via GPR43, synergizing with reduced LPS/TLR4 signaling to quell inflammation.

BA/FXR & LPS/Inflammation: Secondary BAs activating FXR can exert anti-inflammatory effects in the liver and intestine. FXR activation suppresses hepatic inflammation and may contribute to maintaining gut barrier function. Reduced systemic inflammation, in turn, may improve BA metabolism and signaling.

Convergence on Gut-Brain Axis & Central Metabolism: Key metabolites from these axes (SCFAs, secondary BAs, reduced inflammatory signals) collectively signal to the central nervous system via the gut-brain axis (vagal afferents, circulating mediators). They modulate hypothalamic neurons in the arcuate nucleus (ARC) regulating appetite (POMC/CART, AgRP/NPY), energy expenditure, and HPG axis function (kisspeptin, GnRH neurons). This central integration coordinates peripheral metabolic improvements (glucose/lipid handling, insulin sensitivity) with endocrine normalization (reduced androgens, improved gonadotropin dynamics).

This intricate crosstalk ensures that exercise-induced gut microbiota remodeling orchestrates a coordinated multi-organ response, simultaneously targeting metabolic (liver, adipose, muscle), immune (systemic, local gut, ovarian), endocrine (ovary, adrenal, HPG axis), and neural (gut-brain) systems. The resultant metabolic network remodeling—characterized by enhanced insulin sensitivity, optimized hepatic glucose/lipid output, reduced adipose tissue inflammation, suppressed ovarian androgenesis, restored central energy balance, and normalized reproductive endocrine rhythms—collectively reverses the core pathological features of PCOS.

Hyperandrogenism pathologically characterizes an imbalance between increased androgen synthesis and reduced metabolic clearance. Exercise targets the metabolic pathways in the ovaries and liver to restore androgen homeostasis.

(1) Ovarian Targeting to Inhibit Androgen Synthesis

Exercise interventions have been shown to increase butyrate levels. Butyrate, as a histone deacetylase (HDAC) inhibitor (84, 85), has experimental evidence indicating that it can reduce histone H3K9 acetylation at the CYP17A1 promoter in relevant cell types (32), thereby blocking the transcriptional activation of this rate-limiting enzyme in androgen synthesis.

Exercise reduces plasma lipopolysaccharide (LPS) levels and diminishes LPS binding to ovarian Toll-like receptor 4 (TLR4), thereby inhibiting activation of the MyD88/NF-κB signaling axis. Consequently, this reduction alleviates the inhibitory effect of inflammatory cytokines (TNF-α, IL-6) on aromatase (CYP19A1), promoting the conversion of androgens to estrogens. Supporting this mechanism, clinical studies demonstrate that both sprint interval training and moderate-intensity continuous training significantly reduce plasma LPS and TNF-α levels after just two weeks (102).

(2) Liver Targeting to Enhance Androgen Clearance

Exercise induces the production of secondary BAs. These secondary BAs activate hepatic Farnesoid X receptor alpha (FXRα) (103), which upregulates the expression of androgen-metabolizing enzymes SULT2A1 and CYP3A4. This accelerates the sulfation and hydroxylation inactivation of testosterone.

Exercise can modulate the gut-liver axis by increasing beneficial gut bacteria and suppressing the growth of Prevotella, thereby reducing the production of bacterial β-glucuronidase. This enzyme, produced by bacteria such as Prevotella, hydrolyzes conjugated androgens back into their free, active forms. Reducing β-glucuronidase activity is crucial for limiting the enterohepatic circulation and reabsorption of androgens, thereby contributing to the regulation of systemic androgen levels.

Improving IR involves a synergistic interaction between peripheral metabolic regulation and central appetite suppression. Exercise targets key metabolic organs through modulation of the gut-host axis to provide effective intervention.

(1) Peripheral Metabolic Regulation in the Liver and Adipose Tissue

Exercise promotes metabolic homeostasis through the actions of secondary BAs and SCFAs on dual targets: the liver and adipose tissue. In the liver, secondary BAs activate the FXR, which directly inhibits the transcription of genes encoding phosphoenolpyruvate carboxykinase (PEPCK) and glucose-6-phosphatase (G6Pase). These enzymes are key rate-limiting factors in gluconeogenesis and glycogenolysis; their reduced activity significantly attenuates hepatic glucose production, thereby improving hyperglycemia (95). Meanwhile, in adipose tissue, SCFAs interact with the GPR43 receptor, establishing a multi-layered anti-inflammatory barrier: locally, GPR43 activation directly suppresses the assembly of the NLRP3 inflammasome, blocking the release of pro-inflammatory cytokines such as IL-1β and IL-18 (104); systemically, SCFAs regulate the migration threshold of monocytes/neutrophils and the dynamic balance between pro-inflammatory and anti-inflammatory cytokines, thereby curbing the spread of systemic inflammation (105). SCFAs and GPR43 engage in a bidirectional regulatory loop: physiological concentrations of SCFAs upregulate GPR43 expression via histone acetylation (106), and activated GPR43 further inhibits the NF-κB/MAPK signaling axis, reducing NLRP3 inflammasome activity (106). This hepatic-adipose metabolic synergy optimizes hepatic energy output via the FXR-PEPCK/G6Pase axis, while adipose tissue mitigates metabolic inflammation through the GPR43-NLRP3 pathway. Together, they form a dynamic equilibrium system maintaining “glucose-lipid metabolism-inflammation homeostasis.”

(2) Repairing the gut-brain axis to regulate central appetite

Exercise intervention enhances the production of SCFAs and BAs, regulating the gut-brain axis via a cooperative multi-receptor mechanism to maintain energy homeostasis. SCFAs activate free fatty acid receptors FFAR2/GPR43 and FFAR3/GPR41 on gut endocrine L-cells, stimulating the secretion of GLP-1 and peptide YY (PYY) (107–110). Acetate crosses the blood-brain barrier and accumulates in the hypothalamic arcuate nucleus (ARC), where it activates pro-opiomelanocortin (POMC) neurons through an FFAR2-dependent pathway (111), inhibiting agouti-related peptide (AgRP) neurons and enhancing central satiety signals. Butyrate exerts potent appetite suppression by upregulating PYY and inhibiting hypothalamic neuropeptide Y (NPY) neuron activity, largely dependent on vagus nerve-mediated FFAR3 signaling (112–114). Secondary BAs (e.g., DCA, LCA) activate TGR5 highly expressed on vagal afferents and in the ARC, suppressing AgRP/NPY neuron activity (115, 116). The BA-FXR axis further influences central metabolism via the FGF15/19 pathway (117). This integrated signaling dynamically reshapes central appetite regulation and energy homeostasis within the gut-brain axis (118, 119).

Restoring ovulatory function involves re-establishing follicular microenvironment homeostasis and regulating endocrine rhythms. Exercise exerts beneficial effects through dual pathways involving microbiota-derived metabolites and immune modulation.

(1) Repair of the follicular microenvironment

Propionate enhances autophagy by inhibiting the NF-κB and AKT/mTOR signaling pathways and increasing LC3B protein levels. This promotes the clearance of mitochondrial damage induced by trimethylamine N-oxide (TMAO) (120), thereby helping to improve the follicular microenvironment, maintain its homeostasis, and contribute to the amelioration of ovulatory dysfunction.

Immune-Endocrine Crosstalk Regulation. Studies show that gut microbial metabolites metabolize tryptophan into aryl hydrocarbon receptor (AhR) agonists, such as indole-3-lactic acid (ILA), that activate AhR to modulate the host immune response (121). Upon binding ILA, this ligand-dependent transcription factor translocates from the cytoplasm to the nucleus and dimerizes with the AhR nuclear translocator (ARNT) to regulate downstream gene expression. AhR activation promotes the differentiation of regulatory T cells (Treg) (122). Tregs help counterbalance the activity of T helper 17 (Th17) cells, a pro-inflammatory T cell subset whose overactivation can contribute to follicular atresia. By activating AhR, ILA can suppress the differentiation and function of Th17 cells (122), thereby reducing the occurrence of follicular atresia.

(2) Regulation of endocrine rhythms

Normalization of GnRH Pulsatile Secretion. The gut microbiota influences ovarian function partly by modulating the hypothalamic-pituitary-gonadal (HPG) axis. Gut microbiota-derived metabolites, particularly SCFAs and BAs, are potent regulators of hypothalamic GnRH neuron function. Research indicates that alterations in gut microbiota composition (e.g., reductions in Firmicutes and Bacteroidetes) can increase serum levels of GLP-1 (123). GLP-1, produced by intestinal L-cells, stimulates GnRH neurons, potentially involving kisspeptin neuron intermediation. Specifically, the SCFAs (e.g., propionate) and potentially microbial-modulated BAs signal via the gut-brain axis (likely involving vagal afferents and circulating factors) to influence kisspeptin neuron activity in the hypothalamus. Kisspeptin is a critical regulator of GnRH neuron pulsatility. Exercise-induced normalization of this signaling contributes to lowering the LH/FSH ratio and restoring physiological GnRH release patterns.

Enhanced Estrogen Response. Acetate inhibits histone deacetylases (HDACs), helping to maintain normal ovarian function and follicular development, leading to elevated circulating levels of 17β-estradiol (110). Supporting this, animal studies in PCOS models demonstrate that acetate treatment significantly inhibits HDAC activity in plasma and ovaries. This inhibition counteracts the pro-inflammatory effects of TNF-α and restores nuclear factor erythroid 2–related factor 2 (Nrf2) signaling along with related antioxidant defenses (glutathione - GSH, glutathione peroxidase - GPx) (124). Furthermore, acetate administration lowers plasma testosterone levels and the LH/FSH ratio while increasing circulating 17β-estradiol and sex hormone-binding globulin (SHBG) levels in these models (124). Collectively, acetate’s HDAC inhibition contributes to improved ovarian steroidogenesis and estrogen response.

This figure ( Figure 4 ) summarizes the core mechanisms through which exercise-induced changes in the gut microbiota improve the core symptoms of PCOS. These include the reorganization of the SCFAs network, activation of the BA-FXR signaling axis, and inhibition of the LPS-inflammation axis, which collectively target hyperandrogenism, insulin resistance, and ovulatory dysfunction in a tissue-specific manner.

Alterations in Gut Oxygen Supply: Exercise, particularly endurance training, improves cardiovascular function and may temporarily increase gut perfusion and oxygenation (125). This favors the expansion of facultative anaerobes (e.g., certain Lactobacillus spp.) and obligate aerobes, while potentially limiting strictly anaerobic bacteria, including many SCFAs producers. However, the net effect of exercise is generally an increase in SCFAs producers, suggesting the presence of compensatory mechanisms or ecological niche adaptation (78).

BA Flow and Composition: Exercise enhances the excretion of BAs (86, 87) and regulates the gut-liver axis. Changes in the BA profile (e.g., increased secondary BAs) can specifically inhibit or promote the growth of certain bacteria that are sensitive to these antimicrobial molecules (42, 93).

Immune System Regulation: Exercise alters both systemic and mucosal immunity. The release of exercise-induced myokines (e.g., IL-6) and changes in immunoglobulin A (IgA) secretion can influence microbial growth and adhesion (54).

The heterogeneity of PCOS is well-recognized and is formally categorized into phenotypic subtypes based on the Rotterdam diagnostic criteria, which require the presence of at least two of the following three features: (1) hyperandrogenism (clinical and/or biochemical), (2) ovulatory dysfunction, and (3) polycystic ovarian morphology on ultrasound (126). This yields four distinct phenotypes: (1)Phenotype A (classic PCOS): Hyperandrogenism + Ovulatory dysfunction + Polycystic ovaries; (2)Phenotype B: Hyperandrogenism + Ovulatory dysfunction; (3)Phenotype C: Hyperandrogenism + Polycystic ovaries; (4)Phenotype D: Ovulatory dysfunction + Polycystic ovaries (non-hyperandrogenic). Beyond this phenotypic classification, stratification by the presence of obesity (e.g., BMI ≥25 kg/m² vs. lean) and the severity of insulin resistance (e.g., HOMA-IR >2.1 or based on glucose clamp studies) is crucial for understanding metabolic risk and tailoring therapy (127, 128). It is within this framework of phenotypic and metabolic heterogeneity that the differential effects of exercise modalities must be considered.

Given the heterogeneity of polycystic ovary syndrome (PCOS) (e.g., obesity vs. lean, IR-dominant vs. hyperandrogenism-dominant), and the unique physiological effects of different types of exercise, specific interventions for certain subtypes may be more effective. Critically, these exercise modalities engage distinct metabolic pathways, which underpins their differential therapeutic effects.

Endurance/Aerobic Training: This modality primarily enhances cardiovascular health and insulin sensitivity through AMPK activation and improved mitochondrial biogenesis in skeletal muscle. Its significant effects on increasing SCFAs producers (e.g., Faecalibacterium, Roseburia) and reducing systemic inflammation (72, 74, 102) may be especially beneficial for PCOS women with IR and metabolic dysfunction, regardless of obesity status. The potential increase in Akkermansia muciniphila (73) also supports metabolic health.

Resistance/Strength Training: This is effective in increasing muscle mass, improving basal metabolic rate, and glucose handling capacity primarily via mTORC1-mediated muscle protein synthesis. Its documented ability to reduce opportunistic pathogens (e.g., Pseudomonas, Aeromonas) and circulating LPS levels (68, 74) may offer an advantage for PCOS women with significant chronic low-grade inflammation and androgen excess, particularly in the context of obesity, where inflammation is often exacerbated.

HIIT: HIIT provides time-efficient improvements in metabolic health and cardiovascular function by stimulating both aerobic and anaerobic systems, leading to exaggerated post-exercise oxygen consumption (EPOC) and enhanced GLUT4 translocation. Although studies specifically targeting the relationship between PCOS and gut microbiome changes with HIIT are limited, its significant effects on insulin sensitivity and reduction of androgens (76) suggest it may be beneficial for various subtypes, though its intensity must be carefully personalized based on individual tolerance.

Understanding these ecological driving factors and the distinct microbial and metabolic signatures induced by specific exercise patterns (e.g., HIIT’s rapid metabolic improvement linked to specific SCFAs producers; resistance training’s anti-inflammatory effect via pathogen reduction) provides a molecular rationale for designing personalized exercise prescriptions. These prescriptions can be tailored to target the primary pathological drivers in different PCOS phenotypes: Endurance/aerobic training may be prioritized for IR-dominant phenotypes (regardless of BMI) due to its potent effects on SCFAs producers and systemic inflammation. Resistance training might offer advantages for obese PCOS women with significant chronic inflammation and androgen excess by reducing LPS and opportunistic pathogens. HIIT’s efficiency in improving insulin sensitivity and reducing androgens suggests broad applicability, but intensity should be personalized.

Furthermore, the efficacy of exercise is modulated by dietary patterns, creating a critical exercise-diet interaction that shapes the gut microbiota. For instance, the enrichment of SCFA-producing bacteria (e.g., Faecalibacterium) through exercise can be synergistically amplified by a high-fiber, prebiotic-rich diet, which provides essential substrates for microbial fermentation (129). Conversely, the benefits of exercise may be diminished by a Western-style diet high in saturated fats and refined sugars, which promotes dysbiosis and inflammation. Therefore, combining tailored exercise protocols with evidence-based dietary interventions (e.g., Mediterranean or low-glycemic index diets) represents a most promising strategy for optimizing gut microbiome remodeling and achieving sustainable metabolic and endocrine improvements in PCOS (130, 131).

It is noteworthy that the beneficial role of exercise presents a dose-dependent characteristic. This review focuses on the mechanisms by which moderate exercise ameliorates PCOS; however, it is observed in clinical practice that prolonged high-intensity exercise, often coupled with psychological stress (e.g., in some athletes), can conversely lead to endocrine dysfunction and menstrual disturbances. This apparent paradox highlights the critical importance of exercise intensity and individual tolerance. The underlying mechanism may involve excessive activation of the hypothalamic-pituitary-adrenal (HPA) axis, leading to elevated cortisol levels that suppress GnRH pulsatility and ovarian function (132), while also potentially exacerbating gut dysbiosis and inflammation, thereby counteracting the benefits brought by moderate exercise. Consequently, the exercise prescriptions recommended for PCOS management should be tailored to the individual, aiming to achieve the beneficial “eustress” that remodels the gut microbiota and metabolism, while avoiding excessive intensity that leads to detrimental “distress”.