The relationship between vaginal and endometrial microbiota and the outcome of Frozen-Thawed embryo transfer during hormone replacement cycles.

The analysis of 102 vaginal and uterine samples revealed a steep increase in microbial diversity detection from 0 to 20 samples, followed by a plateau after approximately 40 samples (See Fig. 1 ). This Species Accumulation Curve, commonly used in microbiome research, reflects the rate at which new OTUs are discovered with increased sampling, indicating that the majority of microbial diversity was captured within the first 40 samples, with diminishing returns thereafter. Fig. 1 Species Accumulation Curve Note: The horizontal axis represents the number of samples, while the vertical axis represents the number of detected OTUs Species Accumulation Curve Note: The horizontal axis represents the number of samples, while the vertical axis represents the number of detected OTUs DNA was successfully extracted and sequenced from 51 vaginal and 51 uterine samples, achieving a 100% success rate. A total of 7,133,816 raw sequences were obtained, of which 5,753,727 (80.65%) passed quality control and chimera filtering and were retained for downstream analysis. After sequence alignment and OTU clustering, a total of 1,545 OTUs were identified. To illustrate the distribution and overlap of OTUs among the four sample groups—A1 (Lactobacillus-dominant vaginal group), A2 (non-Lactobacillus-dominant vaginal group), B1 (Lactobacillus-dominant uterine group), and B2 (non-Lactobacillus-dominant uterine group)—a Venn diagram was constructed (Fig. 2 ). The diagram shows that 157 OTUs were shared across all four groups, reflecting a core microbiota component. Group A1 had 60 unique OTUs, and Group B2 had 240 unique OTUs. No OTUs were exclusively shared between two or three groups, indicating limited overlap outside the core shared taxa. This visual summary clarifies the composition and group-specific distribution of the identified OTUs and resolves the previous confusion related to OTU counting. Fig. 2 Venn Diagram of OTUs across Four Groups The diagram shows the distribution of unique and shared OTUs among groups A1, A2, B1, and B2. Group A2 and B1 exhibited the highest numbers of unique OTUs (240 and 60, respectively), while 157 OTUs were shared by all four groups Venn Diagram of OTUs across Four Groups The diagram shows the distribution of unique and shared OTUs among groups A1, A2, B1, and B2. Group A2 and B1 exhibited the highest numbers of unique OTUs (240 and 60, respectively), while 157 OTUs were shared by all four groups In the vaginal microbiota, Groups A1 and A2 were both dominated by Lactobacillus, with no significant differences in microbial composition between the two groups (See Fig. 3 A and B). Group B1 exhibited a significantly higher abundance of Lactobacillus (55.78%) compared to Group B2 (10.98%), indicating a dominance of beneficial bacteria in B1. In contrast, Group B2 showed a higher relative abundance of Proteobacteria, along with pathogenic genera such as Acinetobacter (≥ 1%) and Novosphingobium (15.62%). At the phylum level, Firmicutes dominated Group B1 (65.32%), while Group B2 had a lower proportion of Firmicutes (44.30%) and a greater diversity of other phyla, including Bacteroidetes and Actinobacteria (See Fig. 3 C and D). Fig. 3 Microbial Abundance Analysis. A  Bar Chart of Phylum-level Species Distribution in Groups A1 and A2. The horizontal axis represents Groups A1 and A2, the vertical axis represents the composition ratio of species at the phylum level, and the different colors in the figure represent different bacteria genera. Group A1 represents the Lactobacillus -dominant vaginal group, and Group A2 represents the Non- lactobacillus -dominant vaginal group. B  Bar Chart of Genus-level Species Distribution in Groups A1 and A2. The horizontal axis represents Group A1 and Group A2, the vertical axis represents the composition ratio of species at the genus level, and different colors in the figure represent different bacteria genera. Group A1 represents the Lactobacillus -dominant vaginal group, and Group A2 represents the Non- lactobacillus -dominant vaginal group. C  Bar Chart of Phylum-level Species Distribution in Groups B1 and B2. The horizontal axis represents Group B1 and Group B2, and the vertical axis represents the composition ratio of species at the phylum level. Different colors in the figure represent different bacteria genera. Group B1 represents the Lactobacillus -dominant uterine group, and Group B2 represents the Non- lactobacillus -dominant uterine group. D  Bar Chart of Genus-level Species Distribution in Groups B1 and B2. The horizontal axis represents Group B1 and Group B2, and the vertical axis represents the composition ratio of species at the genus level. Different colors in the figure represent different bacteria genera. Group B1 represents the Lactobacillus -dominant uterine group, and Group B2 represents the Non- lactobacillus -dominant uterine group Microbial Abundance Analysis. A  Bar Chart of Phylum-level Species Distribution in Groups A1 and A2. The horizontal axis represents Groups A1 and A2, the vertical axis represents the composition ratio of species at the phylum level, and the different colors in the figure represent different bacteria genera. Group A1 represents the Lactobacillus -dominant vaginal group, and Group A2 represents the Non- lactobacillus -dominant vaginal group. B  Bar Chart of Genus-level Species Distribution in Groups A1 and A2. The horizontal axis represents Group A1 and Group A2, the vertical axis represents the composition ratio of species at the genus level, and different colors in the figure represent different bacteria genera. Group A1 represents the Lactobacillus -dominant vaginal group, and Group A2 represents the Non- lactobacillus -dominant vaginal group. C  Bar Chart of Phylum-level Species Distribution in Groups B1 and B2. The horizontal axis represents Group B1 and Group B2, and the vertical axis represents the composition ratio of species at the phylum level. Different colors in the figure represent different bacteria genera. Group B1 represents the Lactobacillus -dominant uterine group, and Group B2 represents the Non- lactobacillus -dominant uterine group. D  Bar Chart of Genus-level Species Distribution in Groups B1 and B2. The horizontal axis represents Group B1 and Group B2, and the vertical axis represents the composition ratio of species at the genus level. Different colors in the figure represent different bacteria genera. Group B1 represents the Lactobacillus -dominant uterine group, and Group B2 represents the Non- lactobacillus -dominant uterine group In brief, Group B1 had a significantly higher abundance of Lactobacillus, while Group B2 was characterized by increased Proteobacteria and the presence of potential pathogens such as Acinetobacter and Novosphingobium. Conversely, Groups A1 and A2 showed no notable differences. Detailed relative abundances of the dominant bacterial genera (≥ 1%) across all groups are presented in Supplementary Table S2. Alpha diversity analysis revealed that microbial diversity in Group B1 was significantly lower than in Group B2, with a lower Shannon index ( P =0.034) and a higher Simpson index ( P =0.017), showing statistically significant differences. In contrast, no significant differences in alpha diversity were observed between Groups A1 and A2 ( P  > 0.05). (See Fig. 4 A and B). Fig. 4 Inter-group Diversity Differences. A  Shannon and Simpson Indices for Groups B1 and B2. Shannon index (left): two-wilcox, P  = 0.034; Simpson index (right): two-wilcox, P =0.017. Group B1 represents the Lactobacillus -dominant uterine group, and group B2 represents the Non- lactobacillus -dominant uterine group. B  Shannon’s and Simpson’s Indices for Groups A1 and A2. Shannon index (left): two-wilcox, P  = 0.547; Simpson index (right): two-wilcox, P  = 0.639. Group A1 represents the Lactobacillus -dominant vaginal group, and group A2 represents the Non- lactobacillus -dominant vaginal group. C  Principal Coordinate Analysis. D  Anosim Inter-group Diversity Differences. A  Shannon and Simpson Indices for Groups B1 and B2. Shannon index (left): two-wilcox, P  = 0.034; Simpson index (right): two-wilcox, P =0.017. Group B1 represents the Lactobacillus -dominant uterine group, and group B2 represents the Non- lactobacillus -dominant uterine group. B  Shannon’s and Simpson’s Indices for Groups A1 and A2. Shannon index (left): two-wilcox, P  = 0.547; Simpson index (right): two-wilcox, P  = 0.639. Group A1 represents the Lactobacillus -dominant vaginal group, and group A2 represents the Non- lactobacillus -dominant vaginal group. C  Principal Coordinate Analysis. D  Anosim Regarding beta diversity, PCoA analysis showed that the microbial composition in Group B1 was more concentrated, while Group B2 exhibited greater dispersion, indicating higher diversity. Anosim analysis further confirmed that there was a significant structural difference between Group B1 and Group B2 ( R  = 0.3812, P  = 0.001). Conversely, the microbial compositions of Groups A1 and A2 were similar, with no significant structural differences detected ( R =−0.0651, P  = 0.976). (See Fig. 4 C and D for details). To provide a more comprehensive overview of alpha diversity differences, we compared six indices—Shannon, Simpson, PD_whole_tree, Chao1, ACE, and SOBS—between Group A ( Lactobacillus -dominant) and Group B (non- Lactobacillus -dominant). These results are illustrated in Fig. 5 . All six indices showed statistically significant reductions in diversity in Group A ( P  < 0.001), indicating a more uniform but less diverse microbial community in the Lactobacillus -dominant group. Fig. 5 Comparison of alpha diversity indices between Group A ( Lactobacillus -dominant) and Group B (non- Lactobacillus -dominant). A  Shannon index: accounts for both richness and evenness; lower in Group A. B  Simpson index: emphasizes evenness; higher values in Group A indicate dominance by fewer taxa. C  PD_whole_tree: considers phylogenetic diversity; significantly reduced in Group A. D  Chao1 index: estimates total species richness based on rare taxa; lower in Group A. E  ACE (Abundance-based Coverage Estimator): another species richness estimator; also lower in Group A. F  SOBS (Observed species): counts actual observed OTUs; reduced in Group A Statistical comparisons were performed using Wilcoxon rank-sum tests. All indices showed significantly lower values in Group A ( p  < 0.001), indicating a less diverse but more uniform microbial community in the Lactobacillus -dominant group Comparison of alpha diversity indices between Group A ( Lactobacillus -dominant) and Group B (non- Lactobacillus -dominant). A  Shannon index: accounts for both richness and evenness; lower in Group A. B  Simpson index: emphasizes evenness; higher values in Group A indicate dominance by fewer taxa. C  PD_whole_tree: considers phylogenetic diversity; significantly reduced in Group A. D  Chao1 index: estimates total species richness based on rare taxa; lower in Group A. E  ACE (Abundance-based Coverage Estimator): another species richness estimator; also lower in Group A. F  SOBS (Observed species): counts actual observed OTUs; reduced in Group A Statistical comparisons were performed using Wilcoxon rank-sum tests. All indices showed significantly lower values in Group A ( p  < 0.001), indicating a less diverse but more uniform microbial community in the Lactobacillus -dominant group LEfSe analysis revealed that Firmicutes and Lactobacillus were significantly enriched in Group B1, while Proteobacteria, Acinetobacter, and other potential pathogenic microorganisms were predominant in Group B2 (Fig. 6 ). Fig. 6 LDA Results for Differences in Bacterial Populations between Group B1 and Group B2 Note: LDA is Linear Discriminant Analysis. The species with LDA Score greater than the set value (4 by default), represents the species with significant abundance difference between the two groups, and the length of the bar chart represents the influence of different species (LDA Score). Group B1 represents the Lactobacillus-dominant uterine group, and group B2 represents the Non-lactobacillus-dominant uterine group LDA Results for Differences in Bacterial Populations between Group B1 and Group B2 Note: LDA is Linear Discriminant Analysis. The species with LDA Score greater than the set value (4 by default), represents the species with significant abundance difference between the two groups, and the length of the bar chart represents the influence of different species (LDA Score). Group B1 represents the Lactobacillus-dominant uterine group, and group B2 represents the Non-lactobacillus-dominant uterine group The study included 51 patients, with 20 in the LD group and 31 in the LND group. There were no statistically significant differences between the two groups in terms of age, BMI, infertility duration, infertility type, or endometrial thickness on the transfer day (all P  > 0.05), indicating baseline comparability between the groups. See Table 1 . Table 1 Comparison of basic information of FET patients in two groups Item LD NLD F P Number of cases(n) 20 31 Female age (years) 30.05 ± 3.63 30.25 ± 2.92 0.635 0.429 Body Mass Index (kg/m²) 20.87 ± 1.79 20.86 ± 2.17 1.114 0.296 Duration of infertility (years) 2(2, 3) 3(2, 4) 0.327 0.744 Endometrial thickness on transfer day (mm) 10.50 ± 1.64 11.03 ± 1.87 0.377 0.542 Type of infertility (%) 0.008 0.927 Primary infertility 60.00(12/20) 61.29(19/31) Secondary infertility 40.00(8/20) 38.71(12/31) FET represents frozen-thawed embryo transfer, LD represents Lactobacillus -Dominant Group, NLD represents Non- Lactobacillus -Dominant Group, positive number/total number in brackets Comparison of basic information of FET patients in two groups FET represents frozen-thawed embryo transfer, LD represents Lactobacillus -Dominant Group, NLD represents Non- Lactobacillus -Dominant Group, positive number/total number in brackets Compared to the NLD group, the LD group demonstrated a significantly higher clinical pregnancy rate (75% vs. 45.16%, P  = 0.036) and live birth rate (65% vs. 29.03%, P  = 0.011). However, the miscarriage rate showed no statistically significant difference between the two groups( P  = 0.215) (Table 2 ). Table 2 Comparison of pregnancy outcomes between two groups of FET patients [% (n)] Item LD NLD Chi-square value P Clinical pregnancy rate (%) 75.00(15/20) 45.16(14/31) 4.413 0.036 Miscarriage rate (%) 13.33(2/15) 35.71(5/14) 0.215 Live birth rate (%) 65.00(13/20) 29.03(9/31) 6.412 0.011 FET represents frozen-thawed embryo transfer, LD represents Lactobacillus -Dominant Group, NLD represents Non- Lactobacillus -Dominant Group, positive number/total number in brackets Comparison of pregnancy outcomes between two groups of FET patients [% (n)] FET represents frozen-thawed embryo transfer, LD represents Lactobacillus -Dominant Group, NLD represents Non- Lactobacillus -Dominant Group, positive number/total number in brackets

Clinical data from 51 patients undergoing their first frozen-thawed embryo transfer at the Putian University Affiliated Hospital Reproductive Medicine Center between January 2021 and December 2022 were analyzed. Based on endometrial microbiota profiling, patients were stratified into two groups: [ 1 ] Lactobacillus -dominant (LD) group, defined as those exhibiting a relative predominance of Lactobacillus spp. at the genus level within the endometrial microbial community; [ 2 ] Non- Lactobacillus -dominant (NLD) group, characterized by the dominance of other genera over Lactobacillus spp. This classification approach aligns with previous studies employing relative dominance as a grouping criterion [ 7 , 8 ]. In the LD group, the vaginal microbiota were categorized as Group A1 ( Lactobacillus -dominant vaginal group) and the uterine microbiota as Group B1 ( Lactobacillus -dominant uterine group). In the NLD group, the vaginal microbiota were categorized as Group A2 (Non- Lactobacillus -dominant vaginal group) and the uterine microbiota as Group B2(Non- Lactobacillus -dominant uterine group). The group definitions are summarized in Supplementary Table S1 .The study was approved by the Ethics Committee of Putian University Affiliated Hospital(Approval No,202114), and written informed consent was obtained from all the patients prior to the procedure. (1) Age < 35 years [ 9 ]; (2) FSH  1.1 ng/ml [ 10 , 11 ]; (3) At least one high-quality D5/D6 blastocyst (according to the Gardner scoring system: Grade 4 and above, both inner cell mass and trophoblast cells are Grade B or higher); (4) Cervical mucus was investigated within the last three months to rule out chlamydia, ureaplasma, and Neisseria gonorrhoeae infections; vaginal discharge was checked to rule out bacterial vaginosis, vulvovaginal candidiasis, trichomoniasis, and nonspecific vaginitis; no symptoms such as vaginal itching or increased discharge; vaginal secretion was rechecked within 7 days before surgery to exclude related infections; (5) No antibiotics, vaginal suppositories, douching, or sexual intercourse in the month prior to sampling. (1) Chromosomal abnormalities in either partner; (2) Uterine anomalies, intrauterine adhesions, submucosal fibroids, endometrial polyps, or endometrial thickening; (3) Untreated or recurrent hydrosalpinx; (4) Moderate to severe endometriosis; (5) Immunological disorders such as antiphospholipid syndrome; (6) Thyroid dysfunction; (7) Severe systemic diseases. Using a hormone replacement therapy, patients start oral administration of estradiol tablets/estradiol and dydrogesterone tablets (Fematon-red tablets, 2 mg/tablet, from Holland Suwei Pharmaceuticals) at 4–6 mg/d from days 2–3 of the menstrual cycle or after drug withdrawal bleeding. The estradiol dosage is adjusted based on the endometrial thickness, aiming for an endometrial thickness of ≥ 8 mm to induce endometrial transformation. Progesterone vaginal release gel (Cronoton, 90 mg/tube, Merck Serono, Germany) is used at 1 tube/d, with blastocyst transfer occurring on day 6. Post-surgery, patients are given oral estradiol and dydrogesterone tablets (Fenmatong-yellow Tablets, 2 mg/tablet, Holland Suwei Pharmaceuticals) at 6 mg/d, combined with 1 tube/d of progesterone vaginal release gel, continuing support until 14 days post-transfer. Clinical pregnancy is defined by the detection of an intrauterine or ectopic gestational sac on ultrasound scan at 5 weeks post-transplantation; pregnancy outcomes are followed up by phone. Miscarriage is defined as a loss of intrauterine pregnancy before 28 weeks with a fetal weight of less than 1000 g, and live birth is defined as a delivery of a live fetus after 28 weeks. Clinical pregnancy rate = number of clinical pregnancy cycles/number of transfer cycles × 100%, miscarriage rate = number of miscarriage cycles/number of clinical pregnancy cycles× 100%, live birth rate = number of live birth cycles/number of transfer cycles × 100%. To ensure consistency, endometrial samples were collected at the same time point for all patients. The day before endometrial transformation, secretions from the upper third of the vagina are collected, along with a small amount of endometrial tissue during standardized endometrial sampling procedures, which are then placed in sterile dry tubes and stored at −80 °C. To minimize contamination, endometrial samples were collected using sterile, single-use samplers after vaginal cleansing with sterile saline. All collection tools and tubes were DNA-free and sterilized. DNA extraction and PCR were performed in a clean laboratory environment using sterile reagents and consumables. Negative extraction and PCR controls were included during processing, and no bacterial DNA was detected, confirming absence of contamination. The frozen vaginal and uterine samples are transported on dry ice for total DNA extraction, quantitative analysis, polymerase chain reaction (PCR), and sequencing analysis. The sequencing was performed using the Novaseq platform (Illumina Inc, USA). Raw sequencing data are processed through assembly quality control and chimera filtering to produce valid data, which are then analyzed for operational taxonomic units (OTUs) clustering and species classification. Sequences with a similarity greater than 97% are considered as the same OTU. Based on OTU clustering results, each OTU sequence is annotated to obtain corresponding species information and abundance distributions. Analyses of OTU abundance, Alpha and Beta diversity are conducted to ascertain the microbial diversity and abundance within the vagina and uterine cavity. This also allows identification of specific OTUs unique to the Lactobacillus -dominant and non- Lactobacillus -dominant groups. The LEfSe (LDA Effect Size) analysis method is used to identify species with significant differences between groups. The analysis was conducted using SPSS 26.0 statistical software. Quantitative data conforming to a normal distribution are expressed as x̄ ± s and compared between groups using the independent samples t-test; data not conforming to normal distribution are represented as M (P25, P75) and compared using the Mann Whitney U test for two independent samples. Count data are expressed as n (%) and analyzed using the chi-square test or Fisher’s exact test. Sequencing data and analysis were performed using R software (Version 2.15.3) and other software, with the Wilcox rank sum test method applied to analyze differences in species diversity between groups. The LEfSe (LDA Effect Size) analysis was used to identify species with significant differences between the groups and to plot the distribution of LDA values. A P-value < 0.05 was considered statistically significant.

This study utilized 16 S rRNA gene sequencing technology to analyze changes in the reproductive tract microbiota and their impact on pregnancy outcomes in FET patients undergoing hormone replacement therapy. Compared with previous studies, our work focused on a strictly defined first-time FET cohort under hormone replacement cycles, which helps minimize confounding variables such as repeated transfers, previous infections, or hormonal fluctuations. The results showed that a variety of microbial communities were present in both the vagina and endometrium of FET patients. Among them, 88.24% (45/51) of patients had Lactobacillus as the dominant genus in the vagina, while only 39.22% (20/51) of patients had Lactobacillus as the dominant genus in the uterine cavity. The pregnancy rate and live birth rate were higher in the group with Lactobacillus -dominant endometrial microbiota compared to the non- Lactobacillus -dominant group. Most studies indicate that Lactobacillus is the dominant genus in the vaginal microbiota of both fertile and infertile women [ 8 , 12 ], while the proportion of Lactobacillus in the uterine cavity of infertile women varies significantly [ 7 ]. Our findings indicate that Lactobacillus predominates more frequently in the vaginal than in the uterine microbiota, indicating differences between vaginal and uterine microbial communities, which is consistent with previous studies [ 4 ]. Therefore, this study did not group patients based on vaginal microbiota but instead classified FET patients into an LD group ( Lactobacillus -dominant uterine microbiota) and an NLD group because significant structural and compositional differences were only observed in uterine, not vaginal, microbiota.In contrast, the vaginal microbiota (Groups A1 vs. A2) showed no statistically significant differences, indicating limited discriminatory value for outcome prediction. Given that uterine colonization more directly impacts implantation and pregnancy outcomes, uterine microbiota served as the primary basis for grouping. In this study, alpha diversity analysis showed no significant difference in vaginal microbiota diversity between the LD and NLD groups (A1 and A2 groups), but the uterine microbiota diversity was lower in the LD group (B1 group) compared to the NLD group (B2 group). This was further supported by six alpha diversity indices (Shannon, Simpson, PD_whole_tree, Chao1, ACE, SOBS), which were all significantly lower in the Lactobacillus -dominant group, suggesting a simpler and more stable endometrial microbiota potentially favorable for implantation. Beta diversity analysis indicated no significant difference in the vaginal microbiota structure between the two groups, but there was a clear difference in the uterine microbiota structure between the groups. Some studies have shown that as female reproductive tract microbiota ascend, the proportion of Lactobacillus decreases while bacterial diversity increases [ 8 ]. Other studies suggest that the endometrial microbiota is not entirely identical to the vaginal microbiota and that the former is not solely derived from an ascending vaginal infection [ 4 ]. There is also research indicating that microorganisms colonizing the gut and oral cavity may reach the endometrium through hematogenous transmission [ 13 ]. Additionally, various factors such as race, environment, and behavior influence the reproductive tract microbiota. In this study, some FET patients exhibited dysbiosis in the endometrial microbiota without significant differences in their vaginal microbiota structure. Therefore, the stability of vaginal microbiota does not directly reflect the stability of endometrial microbiota, which may be influenced by microbiota from other sites, as well as environmental and behavioral factors. This study’s analysis of the differences in uterine microbiota structure between the LD and NLD groups indicates that the former (B1 group) had significantly increased abundance of Firmicutes and Lactobacillus, while the latter (B2 group) showed higher abundance of Proteobacteria, Bradyrhizobium, Acinetobacter, and Bacillus. Microorganisms can influence normal body functions. Studies have shown that the mechanisms by which Lactobacillus inhibits the growth of pathogenic microorganisms are complex and include creating a mildly acidic environment, secreting hydrogen peroxide and bacteriocins, and participating in immune regulation [ 14 , 15 ]. In addition to these classical pathways, recent studies have explored broader probiotic-related mechanisms in various medical contexts. For example, probiotics have been shown to enhance host immunity and modulate microbial environments relevant to viral infections, systemic inflammation, and reproductive tract health [ 16 – 18 ]. These findings provide important context for the potential therapeutic modulation of the endometrial microbiota to improve ART outcomes. Proteobacteria, as a sign of intestinal flora imbalance, is associated with a variety of intestinal diseases, and its pathogenic mechanisms include dysfunction of intestinal mucosal epithelial cells, inflammation and influence on the host immune system [ 19 ]. Proteobacteria is also associated with extraintestinal diseases. A comparative study on the flora of healthy sows and sows with endometritis found that Proteobacteria showed significant differences between the two groups, with a low abundance in healthy sows but a high abundance in sows with endometritis, indicating that Proteobacteria may induce endometrial lesions [ 20 ]. Proteus outer membrane lipopolysaccharides are result from an inflammatory response and metabolic disorders. Studies have indicated that the expression of pro-inflammatory factors IL-1β, IL-6, IL-8, IL-18 and TNF-α increased after co-culture of lipopolysaccharides with endometrial cells [ 21 ]. Additionally, studies have shown that Clostridium species in the gut can produce toxins that affect the function of the intestinal mucosal epithelium [ 22 ]. To further address the biological relevance of our findings, the mechanisms by which endometrial microbiota influence embryo implantation and pregnancy maintenance are not yet fully understood, but possible mechanisms include: (1) Competitive inhibition, where Lactobacillus, as the dominant genus in the uterine microbiota, inhibits the growth of pathogenic microorganisms by preventing their adhesion to the endometrium; (2) Altering the proportion and function of various immune cells in the endometrium, increasing the release of inflammatory cytokines, and thereby affecting endometrial receptivity; (3) Directly acting on the endometrium and influencing the process of endometrial decidualization.For instance, Lactobacillus spp. produce lactic acid, which maintains a locally acidic environment (pH < 4.5) unfavorable to pathogens, while also contributing to the stabilization of the epithelial barrier. They can secrete antimicrobial substances such as hydrogen peroxide (H₂O₂), bacteriocins, and biosurfactants that inhibit colonization by harmful species [ 14 , 15 ]. In addition, Lactobacillus may modulate the endometrial immune microenvironment by promoting anti-inflammatory cytokines and regulatory T-cell populations, thereby facilitating embryo implantation [ 13 , 15 ]. On the other hand, pathogenic bacteria like Proteobacteria may induce pro-inflammatory signaling via lipopolysaccharides (LPS), leading to impaired receptivity and even early miscarriage [ 19 , 21 ]. These pathways highlight how the microbial environment in the endometrium could directly impact the success of embryo transfer. In this study, there were no statistically significant differences between the two groups in terms of age, body mass index, duration of infertility, type of infertility, and endometrial thickness on the day of transfer. However, the clinical pregnancy rate and live birth rate in the LD group were significantly higher than those in the NLD group, with a statistically significant difference. Although the difference in miscarriage rates between the two groups was not statistically significant, the miscarriage rate in the LD group was noticeably lower than in the NLD group. Some studies have shown that an endometrial microbiota dominated by Lactobacillus is associated with better ART outcomes [ 4 , 7 ]. However, other studies have indicated that Lactobacillus is commonly found in the endometrial microbiota, there is no significant difference in the composition and diversity of microbiota between the pregnant and non-pregnant groups during embryo transfer [ 6 ]. This may be related to different Lactobacillus subspecies in the endometrial microbiota. However, this study has certain limitations.First, as a retrospective cohort study, the strength of the conclusions is inherently limited, although we attempted to exclude major confounding factors that could bias the results.Second, the relatively small sample size may reduce the statistical power to detect differences in secondary outcomes, such as miscarriage rates. Although there was a trend toward lower miscarriage rates in the LD group, this difference did not reach statistical significance. This raises the possibility of a type II error, where a true difference exists but was not detected due to limited power. As this was an exploratory retrospective study, no formal sample size calculation was conducted a priori. However, a species accumulation curve suggested that microbiota diversity reached a plateau after approximately 40 samples, indicating adequate sampling for microbial profiling. Moreover, no formal power calculation was conducted prior to data collection given the exploratory nature of the study. In future prospective studies, we recommend conducting a priori sample size estimations to ensure sufficient power, particularly for detecting differences in less frequent outcomes such as miscarriage.Third, although we controlled for major confounders, certain unmeasured variables—such as long-term dietary patterns, probiotic use, and lifestyle factors—may have influenced the reproductive tract microbiota and were not systematically recorded.Fourth, due to the study design and lack of direct one-to-one pairing between vaginal and uterine samples, we were unable to conduct correlation or network-based analyses to assess microbial continuity between compartments. While we compared alpha diversity across groups, this design limitation restricted our ability to quantitatively evaluate the direct relationship between vaginal and uterine microbiota.Finally, as a single-center exploratory study, our findings require validation in larger, prospective cohorts with longitudinal and matched sampling, which would enable more robust assessments of microbial dynamics and their impact on FET outcomes. In conclusion, having Lactobacillus as the dominant microorganism in the uterine microbiota is beneficial for favorable FET outcomes, while increased diversity in the uterine microbiota, particularly when non-Lactobacillus species dominate, may negatively impact FET outcomes. Vaginal microbiota cannot fully reflect the uterine microbiota. Therefore, even in FET patients with normal vaginal microbiota, attention should also be given to the state of the uterine microbiota. Furthermore, our findings suggest that routine vaginal microbiota evaluation may be insufficient for predicting uterine status, highlighting the need for endometrial-specific assessments in reproductive medicine. In light of these findings, we propose that targeted endometrial microbiota screening and microbiome-modulating interventions, such as probiotics, may be considered in selected clinical scenarios to optimize FET outcomes. These strategies should be explored further in well-designed prospective studies.

With advancements in assisted reproductive technology (ART) and the widespread adoption of vitrification techniques, the number of frozen-thawed embryo transfer (FET) cycles has steadily increased. In FET endometrial preparation, hormone replacement therapy (HRT) protocols are widely used in clinics due to their flexibility and simplicity of operation. Despite advancements in assisted reproductive technologies, clinical pregnancy rates per transfer remain suboptimal (30–40%), likely due to factors such as embryo quality, immune status, and endometrial receptivity [ 1 ]. Recent studies increasingly suggest that changes in the reproductive tract microbiota may affect assisted reproductive outcomes. Eckert et al. [ 2 ] showed that in patients undergoing in vitro fertilization-embryo transfer (IVF-ET), those with vaginal microbiota dysbiosis had significantly reduced pregnancy rates and increased rates of early miscarriage. Another study reported lower live birth rates among first-time FET patients with vaginal microbiota dysbiosis [ 3 ]. Moreno et al. [ 4 ] found that IVF patients with Lactobacillus-dominated microbiota in a receptive endometrium had significantly higher rates of embryo implantation, clinical pregnancy, and live births. However, other studies indicate that the significance of the uterine microbiota to embryo implantation exceeds that of the vaginal microbiota [ 5 ]. Nevertheless, Franasiak et al. [ 6 ] found no significant difference in the composition and diversity of the endometrial microbiota between pregnant and non-pregnant groups in single embryo transfers. Thus, whether reproductive tract microbiota dysbiosis affects assisted reproductive outcomes remains inconclusive. To minimize confounding factors associated with prior embryo transfers—such as prolonged hormone exposure, altered endometrial response, or psychological stress, we included only patients undergoing their first frozen-thawed embryo transfer (FET) cycle. This study is a retrospective analysis of patients undergoing their first FET using a hormone replacement therapy (HRT) method, aimed at exploring the impact of vaginal and uterine microbiota on clinical pregnancy outcomes of FET cycles, to provide rational guidance in the clinic for FET patients.

Supplementary Material 1. Supplementary Material 1.