Comparison of the metabolome of follicular fluid in GnRH agonist versus antagonist protocols during in vitro fertilization cycles.

A total of 137 patients were initially enrolled, including 64 in the GnRH agonist group and 73 in the GnRH antagonist group. Significant differences were observed between the two groups in terms of AMH and AFC ( P  < 0.01). Therefore, propensity score matching was performed to balance age, BMI, AMH, and AFC. After matching, 94 patients remained, with 47 in the GnRH agonist group and 47 in the GnRH antagonist group. The clinical characteristics are summarized in Table  1 . Following matching, no significant differences were noted between the two groups with respect to age, BMI, AMH, AFC, infertility duration, number of previous IVF cycles, or basal sex hormone levels. Table 1 Clinical characteristics of the participants GnRH agonist group ( n  = 47) GnRH antagonist group ( n  = 47) P -value Age (years) 30.8 ± 3.4 31.1 ± 3.0 0.68 c BMI (kg/m 2 ) 21.9 (19.9, 24.3) 21.2 (19.9, 23.2) 0.56 a Infertility duration (years) 3 (2, 5) 2 (2, 5) 0.18 a Previous IVF cycles 1 (1, 1) 1 (1, 1) 0.30 a AMH (ng/ml) 4.6 (3.7, 5.2) 3.9 (2.9, 6.0) 0.38 a AFC 14 (11, 18) 12 (10, 17) 0.25 a Basal FSH (IU/L) 6.7 (5.7, 7.7) 7.0 (5.9, 7.6) 0.59 a Basal LH (IU/L) 4.4 (3.5, 6.1) 4.9 (3.7, 6.4) 0.09 a Basal E2 (pg/ml) 39.4 (31.9, 50.6) 39.3 (32.1, 55.6) 0.96 a Basal P (ng/ml) 0.3 (0.2, 0.4) 0.4 (0.2, 0.6) 0.15 a E2 on the day of hCG (pg/ml) 3424.0 (2526.5, 4066.5) 2714.0 (1896.0, 4559.0) 0.39 a P on the day of hCG (pg/ml) 0.8 (0.5, 1.2) 0.7 (0.5, 1.2) 0.85 a Total Gn dose (IU) 2275 (1800, 2775) 1950 (1650, 2250) 0.03 a No. of retrieved oocytes 13.6 ± 4.1 9.5 ± 3.9 <0.01 c 2PN fertilization rate 68.47% (404/590) 75.07% (286/381) 0.03 b No. of Day 3 embryos 9 (6, 12) 5 (4, 7) <0.01 a No. of Day 3 high-quality embryos 3 (1, 4) 2 (1, 3) <0.01 a Percentage of Day 3 high-quality embryos 32.36% (144/445) 29.97% (92/307) 0.48 b No. of blastocysts 2(1, 4) 1 (0, 2) <0.01 a Blastocyst formation rate 50% (33%. 67%) 50% (0%, 75%) 0.48 a No. of high-quality blastocysts 1 (0, 2) 0 (0, 1) 0.03 a High-quality blastocyst rate 33% (0%, 75%) 0% (0%, 50%) 0.11 a Cumulative clinical pregnancy rate * 72.34% (34/47) 55.32% (26/47) 0.05 b Cumulative live birth rate * 61.70% (29/47) 48.94% (23/47) 0.19 b Abbreviations : BMI Body Mass Index, AFC Antral Follicle Count, hCG human Chorionic Gonadotropin, FSH Follicle-Stimulating Hormone, LH Luteinizing Hormone, E2 Estradiol, P Progesterone, AMH Anti-Müllerian Hormone, Gn Gonadotropin, 2PN Two Pronuclei, IU International Unit a Mann–Whitney U test. The data were expressed as the median (25th percentile, 75th percentile) b Pearson’s chi-squared test. The data were expressed as percentage (numerator/denominator) c Student t-test. The data were expressed as the mean ± SD *Data as of September 2025 Clinical characteristics of the participants Abbreviations : BMI Body Mass Index, AFC Antral Follicle Count, hCG human Chorionic Gonadotropin, FSH Follicle-Stimulating Hormone, LH Luteinizing Hormone, E2 Estradiol, P Progesterone, AMH Anti-Müllerian Hormone, Gn Gonadotropin, 2PN Two Pronuclei, IU International Unit a Mann–Whitney U test. The data were expressed as the median (25th percentile, 75th percentile) b Pearson’s chi-squared test. The data were expressed as percentage (numerator/denominator) c Student t-test. The data were expressed as the mean ± SD *Data as of September 2025 Notably, ovarian response in the GnRH agonist group was superior, reflected in a higher number of retrieved oocytes, which consequently led to a greater number of Day 3 embryos, Day 3 high-quality embryos, blastocysts, and high-quality blastocysts. However, no significant differences were observed between the groups for the percentage of Day 3 high-quality embryos, blastocyst formation rate, high-quality blastocyst rate, or cumulative live birth rate. A total of 111 metabolites were identified in FF samples. Among these, 58 metabolites showed significant differences in abundance between the GnRH agonist and antagonist groups (Table S1, P  < 0.05). OPLS-DA demonstrated a clear metabolic separation between the two protocols (Fig.  1 A), indicating distinct metabolic signatures. Thirteen metabolites were identified as major contributors to this discrimination, with VIP scores greater than 1.5 (Fig.  1 B). Specifically, the concentrations of 11, 14, 17-eicosatrienoic acid, homo-γ-linolenic acid, pentadecanoic acid, and eicosapentaenoic acid were markedly lower in the GnRH antagonist group, with fold changes below 0.75 compared to the GnRH agonist group (Fig.  1 C). Furthermore, 11, 14, 17-eicosatrienoic acid, homo-γ-linolenic acid, and pentadecanoic acid demonstrated excellent diagnostic potential, with area under the curve (AUC) values exceeding 80% (Fig.  2 ). Fig. 1 Metabolite profiling and differential expression analysis of GnRH agonist and antagonist groups. A OPLS-DA score plot depicting a clear separation between the GnRH agonist and antagonist groups. The model’s predictive power (Q2 = 0.42) and the correlation index (R2Y = 0.76) indicate a robust discrimination, according to cross-validation. B VIP plots of the OPLS-DA model, identifying differential metabolites between the two groups (VIP > 1.5). C Volcano plot depicting metabolites with a fold change < 0.75, indicating significant differences in concentration between the GnRH agonist and antagonist groups Metabolite profiling and differential expression analysis of GnRH agonist and antagonist groups. A OPLS-DA score plot depicting a clear separation between the GnRH agonist and antagonist groups. The model’s predictive power (Q2 = 0.42) and the correlation index (R2Y = 0.76) indicate a robust discrimination, according to cross-validation. B VIP plots of the OPLS-DA model, identifying differential metabolites between the two groups (VIP > 1.5). C Volcano plot depicting metabolites with a fold change < 0.75, indicating significant differences in concentration between the GnRH agonist and antagonist groups Fig. 2 Bar plots demonstrating the levels of representative metabolites show pronounced differences between the GnRH agonist and antagonist groups (right panels). Receiver operating characteristic (ROC) curves for the representative metabolites exhibit an AUC exceeding 80% (left panels). These ROC curves provide information on the optimal cut-off values, as well as sensitivity and specificity values Bar plots demonstrating the levels of representative metabolites show pronounced differences between the GnRH agonist and antagonist groups (right panels). Receiver operating characteristic (ROC) curves for the representative metabolites exhibit an AUC exceeding 80% (left panels). These ROC curves provide information on the optimal cut-off values, as well as sensitivity and specificity values The number of Day 3 high-quality embryos was positively correlated with 14 metabolites. Crucially, all of these 14 metabolites were among the differentially expressed metabolites that distinguished the GnRH agonist group from the antagonist group, including 11, 14, 17-eicosatrienoic acid, pentadecanoic acid, palmitelaidic acid, 10-heptadecenoic acid, 11, 14-eicosadienoic acid, stearic acid, myristoleic acid, and 2,4-di-tert-butylphenol (Fig.  3 ). Fig. 3 Correlation between the number of Day 3 high-quality embryos and differential metabolite levels. Statistical analyses were conducted using Pearson correlation Correlation between the number of Day 3 high-quality embryos and differential metabolite levels. Statistical analyses were conducted using Pearson correlation KEGG pathway enrichment analysis of the differential metabolites identified “Biosynthesis of unsaturated fatty acids” as the most significantly enriched pathway. Additional significantly enriched pathways included “Linoleic acid metabolism”, “Glutathione metabolism”, and pathways related to amino acid metabolism such as “Glycine, serine and threonine metabolism” and “Alanine, aspartate and glutamate metabolism”.

FF samples were collected from patients undergoing IVF treatment at the Center for Reproductive Medicine, the Second Affiliated Hospital of Chongqing Medical University. Consecutive IVF cycles stimulated with either the GnRH agonist long protocol or the GnRH antagonist protocol between May 2019 and November 2019 were screened for eligibility. Patients were excluded if they met any of the following criteria: (1) BMI ≥ 24 kg/m²; (2) a history of endometriosis, ovarian surgery, endocrine or autoimmune disorders; (3) diagnosis of polycystic ovary syndrome (PCOS); or (4) more than two previous IVF attempts. AFC and baseline sex hormone levels were assessed on menstrual cycle days 2–4. In routine clinical practice at our center, all patients receive standardized lifestyle and dietary counseling—emphasizing a balanced diet and moderate physical activity—approximately 2–3 months prior to cycle initiation to minimize metabolic variability. To further reduce baseline confounding factors associated with age, BMI, AMH, and AFC, propensity score matching was applied in accordance with established methodology [ 11 ]. For patients undergoing the GnRH agonist long protocol, pituitary downregulation was initiated with daily subcutaneous injections of triptorelin (0.1 mg/day; Ferring Pharmaceuticals) during the mid-luteal phase of the preceding menstrual cycle. After 14 days of treatment and confirmation of adequate pituitary suppression, ovarian stimulation commenced using recombinant FSH (rFSH; Gonal-F, Merck Serono; 150–225 IU/day), with dose adjustments guided by transvaginal ultrasound monitoring and serial serum estradiol assessments. For patients in the GnRH antagonist protocol, ovarian stimulation with rFSH was started on menstrual cycle days 2–3. A GnRH antagonist (cetrorelix 0.25 mg/day; Merck Serono) was added once the leading follicle reached ≥ 12 mm in diameter or serum estradiol exceeded 300 pg/mL, and continued daily until the ovulation trigger. Final oocyte maturation was induced using either 6000-10000 IU urinary hCG (Lizhu Pharmaceutical Trading, China) or 250 µg recombinant hCG (Merck Serono, Italy) when at least two follicles measured > 18 mm. In GnRH antagonist cycles, patients having a high risk of OHSS received a GnRH agonist trigger (0.2 mg) alone or in combination with hCG. Oocyte retrieval was performed 36 h or 37.5 h after triggering under transvaginal ultrasound guidance [ 12 ]. FF samples were aspirated from the first dominant follicle (≥ 18 mm) into pre-chilled sterile tubes. After immediate centrifugation at 1,000 × g for 10 min at 4 °C to remove cellular debris, the supernatant was aliquoted in 300 µl portions and stored at − 80 °C until analysis. For metabolite extraction, 250 µl thawed FF was transferred into 1.5 ml tubes, spiked with 20 µl internal standard (2,3,3,3-d4-alanine, 10 mM), and mixed with 730 µl cold methanol for protein precipitation. Following 40 min of incubation at − 20 °C, samples were centrifuged at 1,500 × g for 20 min, and supernatants were dried in a SpeedVac concentrator (LABCONCO, #7810041) at 1,000 × g for 8 h. Dried extracts were stored at − 80 °C prior to derivatization [ 13 ]. Metabolites extracted from follicular fluid were derivatized using a methyl chloroformate (MCF)-based protocol adapted from previously validated methods [ 13 , 14 ]. Briefly, each dried metabolite extract was reconstituted in 200 µl of 1 M sodium hydroxide. Subsequently, 68 µl of pyridine and 30 µl of MCF were added sequentially, and the mixture was vortexed vigorously for 30 s to facilitate derivatization, thereby reducing the boiling points of target analytes for GC-MS compatibility. A mixture of 300 µl chloroform and 800 µl 50 mM sodium bicarbonate solution was then added and shaken for 10 s to promote phase separation. Following centrifugation at 1, 500 × g for 10 min, the lower chloroform phase containing the derivatized metabolites was carefully collected for GC-MS analysis. To ensure data reliability, all FF samples were analyzed in a single continuous GC-MS run using a randomized injection sequence. Quality control measures included periodic injections of pooled quality control samples and solvent blanks to monitor instrumental stability. After calibration, analyses were conducted on an Agilent 7890B gas chromatograph coupled to an Agilent 5977 A mass selective detector. Metabolite separation was achieved on a ZB-1701 capillary column (30 m × 250 μm i.d. × 0.15 μm film thickness; Phenomenex), equipped with a 5m integrated guard column. The mass detector operated in electron impact ionization mode at 70 eV. GC temperature programming and MS parameters followed previously optimized protocols [ 15 ]. GC-MS analysis of derivatized metabolites was performed using an Agilent Intuvo 7890B gas chromatography system coupled to an Agilent 5977 A mass selective detector (MSD), operated in electron impact ionization mode at 70 eV, following established procedures [ 16 ]. Samples were injected in pulsed splitless mode, with the inlet temperature set at 290 °C. Helium was used as the carrier gas at a constant flow rate of 1.0 mL/min. Chromatographic separation was achieved using a BD-1701 capillary column (30 m × 250 μm i.d. × 0.25 μm film thickness). The oven temperature program began with an isothermal hold at 45 °C, followed by a stepwise ramp to a final temperature of 280 °C, ensuring optimal separation of derivatized metabolites. Raw GC-MS data were processed using the Automated Mass Spectral Deconvolution and Identification System (AMDIS) to generate deconvoluted mass spectra. Preliminary metabolite identifications were made by matching fragmentation patterns and retention times against an in-house spectral library constructed from authenticated reference standards. Tentative identifications were subsequently verified through comparison with the NIST mass spectral database to ensure high confidence in metabolite annotation. Defined as embryos derived from normal fertilization (2PN) that exhibit 7–9 blastomeres at 68 ± 1 h post-insemination, < 10% fragmentation, stage-specific symmetrical blastomeres, absence of multinucleation, and no significant cytoplasmic vacuolization. Defined as blastocysts grading AA/AB or BA/BB and expansion grade 3 or above. Defined as the number of patients who achieved a clinical pregnancy, confirmed by the presence of a gestational sac on ultrasound after any fresh or frozen-thawed embryo transfer originating from a single oocyte retrieval cycle, divided by the total number of patients who underwent oocyte retrieval in that cycle. Defined as the number of patients who achieved at least one live birth (delivery of a viable infant at ≥ 28 weeks of gestation) after any fresh or frozen-thawed embryo transfer originating from a single oocyte retrieval cycle, divided by the total number of patients who underwent oocyte retrieval in that cycle. All data are presented as mean ± standard deviation (SD) or median (25th–75th percentile), as appropriate. Normality and homogeneity of variance were assessed using the Shapiro–Wilk test. Between-group comparisons of clinical characteristics were performed using the Student’s t-test or the Mann–Whitney U test for continuous variables, and Pearson’s chi-square test for categorical variables. A two-sided P-value < 0.05 was considered statistically significant. Given the exploratory nature of the metabolomic analysis, P-values for differential metabolites were not adjusted for multiple comparisons to avoid excessive Type II errors. Instead, the robustness of metabolite selection was ensured by integrating unadjusted P-values with fold-change thresholds (FC) and variable importance in projection (VIP) scores derived from the Orthogonal partial least squares–discriminant analysis (OPLS-DA) model. OPLS-DA modeling and metabolic pathway enrichment were performed using MetaboAnalyst 6.0 ( https://www.metaboanalyst.ca/ ).

This study systematically reveals distinct FF metabolic profiles induced by GnRH agonist and antagonist protocols through GC-MS metabolomics. After rigorously controlling for confounders such as age, BMI, AMH, and AFC via propensity score matching [ 11 ], we identified 58 differentially expressed metabolites. Pathway enrichment analysis indicated that the GnRH agonist protocol was associated with a significant upregulation in the “Biosynthesis of unsaturated fatty acids” pathway and other pathway related to amino acid metabolism (Fig. 4 ). This discovery provides a new metabolic perspective for understanding how different GnRH analogues may influence the follicular microenvironment. Fig. 4 Pathway enrichment analysis of differential metabolites between GnRH agonist and antagonist groups Pathway enrichment analysis of differential metabolites between GnRH agonist and antagonist groups Consistent with the observed superior ovarian response in the GnRH agonist group (higher oocyte yield), which aligns with some recent studies [ 5 , 16 ], we found significantly increased levels of key lipids such as 11,14,17-eicosatrienoic acid, homo-γ-linolenic acid, and pentadecanoic acid. These findings suggest that GnRH analogues may differentially affect follicular biochemical processes via their distinct pharmacological actions on the hypothalamic-pituitary-ovarian axis. The GnRH agonist long protocol, characterized by sustained pituitary desensitization and a potentially more stable FSH environment, might promote a lipid-rich FF profile. In contrast, the immediate LH suppression achieved by GnRH antagonists could theoretically influence theca cell steroidogenesis and the subsequent availability of lipid precursors for estrogen synthesis in granulosa cells [ 17 – 19 ]. The reduction of multiple beneficial lipids in the GnRH antagonist group, including PUFAs and pentadecanoic acid (C15:0), may influence follicular physiology through several biologically plausible avenues. First, as precursors for prostaglandin synthesis, the reduction of omega-3 and omega-6 PUFAs (e.g., eicosapentaenoic acid, EPA) could potentially influence prostaglandin E2–mediated expansion of the cumulus–oocyte complex, a process critical for ovulation and fertilization [ 17 ]. Second, the altered levels of PUFAs (essential components of cell membrane phospholipids) might disrupt membrane fluidity and the efficiency of signal transduction during oocyte activation [ 20 ]. Furthermore, our study is among the first to report a significant reduction of pentadecanoic acid (C15:0) in the GnRH antagonist group. This odd-chain saturated fatty acid, a biomarker of dairy intake, has been implicated in modulating mitochondrial β-oxidation and PPARγ activity [ 21 ]. Its relative deficiency might, therefore, be linked to altered energy metabolism within the follicle, potentially rendering oocytes more susceptible to oxidative stress [ 18 , 22 ]. These mechanistic interpretations are consistent with findings from previous studies. For instance, dietary interventions with omega-3 fatty acids have been shown to prolong reproductive lifespan and improve oocyte quality in model systems [ 23 ]. Moreover, our findings provide a potential metabolic rationale for the hypothesis that omega-3 fatty acid supplementation may improve treatment outcomes among women undergoing IVF [ 24 ]. Collectively, this growing body of evidence, as summarized in the systematic review [ 25 ], underscores the crucial role of lipid metabolism in oocyte competence. Our data specifically suggest that the GnRH antagonist protocol is associated with lower endogenous levels of these beneficial lipids within the follicular microenvironment. However, the proposed associations are hypothetical, and direct causal links require further experimental validation. In the field of FF metabolomics, different analytical platforms reveal distinct metabolic profiles with their own strengths and limitations. This study employed GC-MS, a platform known for its high sensitivity and specificity in quantifying small molecular weight metabolites such as free fatty acids and organic acids. Liquid Chromatography-Tandem Mass Spectrometry (LC-MS/MS) enables broader profiling of complex lipid classes, including phospholipids, thereby revealing alterations in higher-order lipid metabolic networks [ 7 ].Nuclear magnetic resonance (NMR) spectroscopy, owing to its high reproducibility, excels in the systematic analysis of polar metabolites such as amino acids and carbohydrates (e.g., glucose, lactate) [ 28 , 29 ]. However, no studies to date have employed LC-MS/MS or NMR to compare FF metabolomic profiles between different COS protocolsFuture research integrating complementary multi-platform analytical approaches holds promise for a more comprehensive elucidation of how different ovarian stimulation protocols influence the biochemical microenvironment of the ovarian follicle. Beyond elucidating the metabolic distinctions between protocols, our findings hold significant promise for clinical practice. The GnRH antagonist protocol has gained widespread global adoption due to its simplicity, shorter duration, and markedly reduced risk of OHSS [ 12 , 34 , 35 ]. However, questions have persisted regarding its efficacy in yielding oocytes of optimal developmental competence compared to the GnRH agonist long protocol [ 6 , 36 , 37 ]. The identification of specific fatty acid differences in GnRH antagonist cycles provides a metabolic explanation for the decreased normal fertilization rate. More importantly, it unveils a potential therapeutic target. If future interventional studies can confirm that strategic supplementation (e.g., through tailored lipid-enriched culture media or pre-treatment nutritional interventions) can rectify this metabolic gap, it would directly address the current limitation. Such a strategy could harness the superior safety and convenience of the GnRH antagonist protocol while empowering it with the metabolic profile associated with high oocyte quality, without altering its established clinical advantages. The translational potential of these findings lies in identifying actionable targets for improving GnRH antagonist protocol outcomes. Our data suggest that the follicular microenvironment in GnRH antagonist cycles is characterized by a relative deficiency of specific beneficial fatty acids. This provides a reasonable metabolic basis for considering pre-conception or peri-stimulation nutritional interventions. In particular, omega-3 fatty acid supplementation emerges as a strategically supported candidate. Given that metabolites like EPA and its precursors reduced in the GnRH antagonist group, and considering the established role of omega-3 PUFAs in reducing inflammation, improving cell membrane fluidity, and supporting cumulus cell function [ 17 , 23 ], targeted supplementation could be a logical approach to amend this metabolic gap. Although our observational study cannot prove efficacy, it generates a strong hypothesis that future randomized controlled trials should test: whether rectifying this specific lipid signature through omega-3 supplementation can modify follicular metabolic profiles in GnRH antagonist cycles, thereby harnessing the protocol’s practical benefits without compromising oocyte quality. However, our study has several limitations that should be acknowledged. First, our GC-MS approach has a blind spot for many non-volatile and complex lipids (e.g., phospholipids, sphingolipids) that are better captured by LC-MS platforms [ 7 , 10 , 19 ]. A multi-platform approach in future studies would provide a more comprehensive lipidomic profile. Second, the generalizability of our findings is constrained by the exclusion of patients with BMI ≥ 24 kg/m², PCOS, or endometriosis. As obesity and inflammatory conditions significantly alter lipid metabolism [ 32 , 38 ], the external validity of our results to these prevalent patient populations remains to be determined. Third, despite propensity score matching, the observational nature of this study cannot establish causality, and residual confounding from unmeasured factors (e.g., subtle dietary variations, gut microbiome-derived metabolites) cannot be entirely ruled out. Fourth, the sample size, though adequate for an initial discovery-phase investigation, limits the statistical power for detecting subtle effects and for robust subgroup analyses. Furthermore, in this exploratory study, we prioritized the reduction of Type II errors over Type I errors by not adjusting P-values for multiple comparisons [ 39 ]. To counterbalance this approach and ensure robust identification of key metabolites, we relied on a combination of stringent criteria (fold change, VIP scores, and AUC values), which converged to highlight specific fatty acids of interest. In conclusion, this study underscores distinct FF metabolic profiles between GnRH agonist and antagonist protocols, with the GnRH agonist protocol enriching a specific set of fatty acids associated with a superior numerical embryo yield. The identification of specific metabolites correlated with high-quality embryos offers valuable insights into personalizing fertility treatments and evaluating the predictive value of FF metabolomics [ 13 , 31 , 33 , 39 ]. However, these findings must be interpreted with caution considering the limitations inherent in the single-center observational design and the GC-MS platform’s limited coverage of complex lipids. Further studies are required to validate these findings in larger, multi-center cohorts using complementary analytical platforms, and to determine whether modulating these metabolic pathways has the potential to improve outcomes in GnRH antagonist cycles.

Infertility represents a significant public health challenge for reproductive-aged couples worldwide. According to the World Health Organization, it affects approximately 17.5% of the adult population [ 1 ]. In vitro fertilization-embryo transfer (IVF-ET), as a core assisted reproductive technology, has seen rapidly expanding application globally. The number of reported IVF cycles has increased substantially from approximately 1 million in 2010 to about 3 million during 2017–2018, with a continuing upward trend [ 2 ]. The success of IVF largely depends on the retrieval of high-quality oocytes, which is highly reliant on the homeostasis of the intrafollicular microenvironment. Within this environment, the developmental potential of oocytes is determined through intricate bidirectional metabolic crosstalk with surrounding somatic cells [ 3 ]. Follicular fluid (FF), serving as a biochemical reflection of this microenvironment, acts as a “metabolic repository” indicative of oocyte health. The metabolic profile of FF, particularly its lipid and energy metabolism networks, has emerged as a powerful tool for predicting IVF outcomes [ 4 ]. During controlled ovarian stimulation (COS) in IVF, gonadotropin-releasing hormone (GnRH) agonist and antagonist protocols represent two fundamental strategies. Substantial clinical evidence confirms that these protocols exhibit distinct characteristics in key outcomes: the GnRH agonist protocol is often associated with a higher number of retrieved oocytes and potentially superior embryo development, whereas the GnRH antagonist protocol has a clear advantage in significantly reducing the risk of ovarian hyperstimulation syndrome (OHSS), offering a shorter treatment duration, and requiring a lower total gonadotropin dose. Furthermore, the antagonist protocol provides greater flexibility and is favored for patients with a high risk of OHSS. However, the precise biological mechanisms underlying these outcome disparities, particularly at the level of the follicular microenvironment, remain incompletely elucidated [ 5 , 6 ]. In recent years, research has begun to explore the biological basis of these differences through analyzing the composition of FF. A pivotal lipidomic study by Jiang et al. provided direct evidence that GnRH agonist and antagonist protocols elicit distinct FF microenvironments, implicating differential regulation of lipid metabolism and ovarian steroidogenesis [ 7 ]. Beyond this, other metabolomic investigations have demonstrated the general utility of high-coverage metabolomics in capturing FF alterations associated with various reproductive states, thereby underscoring the sensitivity of the FF metabolome to the changeable physiological conditions [ 8 – 10 ]. Although these preliminary findings suggest that protocol-specific metabolic signatures may exist, the broader body of existing research suffers from notable limitations. Critically, most studies have not been rigorously controlled for inherent baseline differences in patient populations (e.g. age, BMI, AMH, AFC), which severely compromises the attribution of any observed metabolic differences to the stimulation protocol itself. Furthermore, a systematic correlation analysis between these putative metabolic markers and key laboratory indicators of embryonic development (e.g. high-quality embryo rate, blastocyst formation rate) or ultimate clinical outcomes is generally lacking. This knowledge gap hinders a deeper understanding of the molecular mechanisms through which different COS protocols influence oocyte quality at the metabolic level and impedes the development of truly individualized COS strategies. Therefore, this study employs gas chromatography-mass spectrometry (GC-MS) to systematically compare FF metabolic profiles between GnRH agonist and antagonist protocols, leveraging the platform’s high sensitivity and strong performance in the relative quantification of key metabolic classes (including organic acids and free fatty acids) to elucidate the underlying metabolic mechanisms.

Supplementary Material 1. Table S1. The peak intensities and P -values of the metabolites showing significantly different levels between GnRH agonist and antagonist groups. Supplementary Material 1. Table S1. The peak intensities and P -values of the metabolites showing significantly different levels between GnRH agonist and antagonist groups.