Gut microbiota changes in patients with hypertensive disorders and gestational diabetes in early pregnancy: a prospective cohort study

Abstract

Background

The potential impact of metabolic diseases during pregnancy is of increasing interest. This prospective cohort study aimed to analyze the gut microbiota changes of pregnant women with gestational diabetes mellitus (GDM) and hypertensive disorders in pregnancy (HDP) before the 20th week of gestation.

Methods

Feces were collected before 20 weeks of pregnancy, where the microbiota was analyzed for using 16S rRNA gene sequencing. QIIME2 and R packages (v3.2.0) were used for sequence data analyses. The diversity and composition of the microbiota of GDM, HDP and healthy pregnancy groups were further compared by following up the pregnant outcomes. A logistic regression model was applied to assess the diagnostical value of gut microbiota for GDM and HDP.

Result

Compared with healthy pregnancies, HDP group showed a higher abundance of Actinobacteria, and a lower abundance of Bacteroidetes at the phylum level (p < 0.05). In contrast, Bacteroidetes abundance was increased in GDM group. Otherwise, Synergistes showed a higher abundance in both GDM and HDP group at the family level. The further results displayed that the abundance level of Actinobacteria may maintained important implications for the adverse development of GDM and HDP women. Subsequently, the logistic regression model showed that the area under the curve for HDP and GDM groups were 0.77 and 0.69, respectively.

Conclusion

Altered gut microbiota before 20 weeks of gestation precedes clinical signs of GDM and HDP, in which the species and composition of Actinobacteria, Bacteroidetes, and Synergistes may be associated with clinical symptoms of the metabolic diseases in pregnancy. This study provides new research strategies to focus on the changes in gut microbiota of early pregnant women.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12884-025-07903-7.

Introduction

Gestational diabetes mellitus (GDM) and hypertensive disorders in pregnancy (HDP) are specific metabolic diseases that occur during pregnancy, with the incidence rates of approximately 20% and 10%, respectively [1]. With the implementation of the two-child policy in China, the proportion of older pregnant women is expected to rise, potentially leading to higher incidence rates. These conditions, characterized by rapid symptom changes, not only act as primary causes of multiple organ damage, obstetric bleeding, and preterm birth in pregnant women, but also as high-risk factors for permanent cardiovascular damage and decreased physical fitness in both the women and their offspring [2–5].

The gut microbiota consists of various types of microorganisms such as bacteria, phages, viruses, fungi and archaea in the intestinal tract, which is a complex and dynamic ecosystem that has a symbiotic relationship with the human body. Human gut microbiota plays an important role in host metabolism, nutrition, immunity and other processes [6]. Previous study has indicated that a correlation between gut microbiota and its metabolites play an essential role in the development of chronic inflammatory diseases, metabolic syndrome, and pregnancy disorders [7]. More than 50 chronic diseases, including obesity, type 2 diabetes, tumors, and autism, have been found correlated with intestinal microbiota in publications [8, 9]. Recent researches report that the gut flora of patients with GDM and HDP significantly differs from that of healthy pregnant women. Chen X et al. found that the gut microbiota of preeclampsia patients could differed from that of women with normal pregnancies in terms of β-diversity, and that the α-diversity of the gut flora of preeclampsia patients was reduce [10]. Jin J et al. discovered that preeclampsia patients had a reduced abundance of short-chain fatty acid-producing bacteria in Akkermansia, Oscillibacter, and that fecal and serum levels of acetic, propionic, butyric, and valeric acids were significantly reduced [11]. Koren O et al. studied 91 pregnant women with varying pre-pregnancy BMI and gestational diabetes statuses. They collected fecal samples from early (T1) to late (T3) pregnancy and found significant differences in gut flora composition between these stages, with α-diversity lower in T3. There was an increase in Proteobacteria and Actinobacteria and a decrease in butyric acid-producing bacteria like Faecalibacterium from T1 to T3 [12]. The latest research on the metagenome of intestinal microbiota during pregnancy has confirmed that there are significant differences in intestinal microbiota from T1 to T3 [12]. This is mainly characterized by an increase in the abundance of Proteobacteria and Actinobacteria, while the overall richness of the community decreases, indicating a markedly decrease in intestinal microbiota diversity during pregnancy.

Whether those differences in the gut microbiota structure before 20 weeks of pregnancy are closely related to gestational hypertension and gestational diabetes, which can effectively predict the occurrence of metabolic disorders during pregnancy, needs to be further investigated. This study aims to establish a predictive model for gestational metabolic disorders based on the differences in the gut microbiota before 20 weeks of pregnancy in the population. It represents significant value for the health management and nutritional regulation of pregnant women in the early and middle stages of pregnancy. Furthermore, it provides new predictive strategy for the early prevention and risk prediction of GDM and HDP, which have profound implications for reducing the incidence of maternal and infant diseases and the prevention and control of adult cardiovascular diseases.

Materials and methods

This study constructed a prospective cohort study to collect clinical data (e.g., age, body mass index, diet) and biological samples throughout pregnancy. Fecal samples will be obtained prior to the 20th week of pregnancy for clinical analysis and sequencing. Screening for HDP and GDM will be performed monthly from the 20th week of pregnancy until its conclusion, with final pregnancy outcomes duly recorded. These fecal samples will undergo 16S rRNA gene sequencing. The flowchart of the experimental design was as shown in Fig. 1.

Fig. 1.

The flowchart of the experimental design

Inclusion and exclusion criteria

The 2020 Practice Bulletin Summary (Number 222, Gestational Hypertension and Preeclampsia) by the American College of Obstetrics and Gynecologist [13]: HDP was defined as a multisystem disorder by the sudden onset of hypertension (systolic blood pressure (SBP) ≥ 140 mmHg and/or diastolic blood pressure (DBP) ≥ 90 mmHg) and proteinuria (> 300 mg/24 h), happening after 20 weeks of gestation in a woman who was previously normotensive. Moreover, it includes new-onset hypertension accompanied by the onset of conditions such as thrombocytopenia (platelet count ≤ 100 × 10⁹/L), renal insufficiency (serum creatinine concentrations ≥ 1.1 mg/dL or a doubling of the serum creatinine concentration in the absence of other renal diseases), impaired liver function (elevated blood concentrations of liver transaminases to twice the normal concentration), pulmonary edema, or a new-onset headache that is unresponsive to medication.

The inclusion criteria for the cohort: (1) pregnant women aged between 20 and 35 years, of normal weight, and within their first trimester (T1); (2) individuals capable of providing informed consent; and (3) participants willing and able to provide fecal samples.

Exclusion criteria include: (1) use of certain medications in the past six months, including systemic antibiotics, corticosteroids, cytokines, methotrexate or other immunosuppressive drugs, high doses of probiotics, hormonal contraceptives, and regular dietary intake of fermented products; (2) testing positive for HIV, HBV, or HCV, as confirmed by immunoblot or molecular testing; (3) symptoms of immune suppression or deficiency, including HIV infection; (4) gastrointestinal surgery, cholecystectomy, or appendectomy within the past five years, excluding bowel resection at any time; (5) chronic diarrhea caused by Clostridium difficile or unknown pathogens, or chronic constipation; (6) pregnancy complications such as miscarriage, preterm birth, fetal malformation, choriocarcinoma; (7) high-risk pregnancy complications like pre-eclampsia, diabetes mellitus, gestational diabetes mellitus; (8) alcohol consumers; (9) Pre-pregnancy comorbidities including obesity (BMI ≥ 30 kg/m²), cardiovascular disease, hepatic dysfunction, or renal disease. To further reduce confounding, gestational weight gain (GWG) was reviewed for all participants up to the 20th week of gestation. Only those whose GWG fell within the Institute of Medicine (IOM)–recommended ranges for their respective pre-pregnancy BMI categories were included in the final analysis.

Screening for GDM: In individuals with no pre-existing diabetes, a diagnosis of GDM is made if diabetes develops anew after 20 weeks of gestation, as determined through an oral glucose tolerance test (OGTT) [14].

Biological sample collection

Provide a sterile bedpan and a fecal collection bag, which includes sterile fecal collection containers, a collection spoon, and toilet paper. The participant should first empty their bladder, then position the sterile bedpan on the toilet, lining it with four sheets of clean toilet paper. The participant should then defecate onto the toilet paper. Following this, a sample of feces from the latter part of the stool should be collected using the spoon and sequentially placed into the three collection containers, filling each to approximately one-third of its volume. The containers should then be sealed, placed into the fecal collection bag, and stored in a −80℃ freezer within 30 min.

16S rRNA gene amplicon sequencing

We used a 30:30:30 ratio for HDP: GDM: control groups. DNA samples were extracted using QIAamp Fast DNA Stool Mini KIT (QIAGEN, Hulsterweg, Netherlands, Netherlands) and evaluated with NanoDrop NC2000 (Thermo Fisher Scientific, Waltham, MA, USA) and agarose gel electrophoresis. The 16S rRNA gene V3-V4 region was amplified using specific primers (forward primer (5’-ACTCCTACGGAGGCAGCA-3’) and the reverse primer (5’-GGACTACHVGGGTWTCTAAT-3’)) and PCR mixture [15]. Multiplex sequencing was facilitated by incorporating sample-specific barcodes. PCR amplicons were purified using Vazyme VAHTS DNA Clean Beads (Vazyme, Nanjing, China) and quantified with the Quant-iT PicoGreen dsDNA Assay Kit (Invitrogen, Carlsbad, CA, USA). After pooling, amplicons were sequenced on the Illumina NovaSeq platform using the NovaSeq 6000 SP Reagent Kit at Magigene Technology Co., Ltd (Guangzhou, China).

Sequence analysis

Microbiome data analysis utilized QIIME2 2019.4 with slight modifications from the official tutorials (Using QIIME to analyze 16S rRNA gene sequences from microbial communities). Initially, sequences were demultiplexed and primers were removed. Next, we underwent quality checks, denoising, merging, and chimera elimination using DADA2 (DADA2: High-resolution sample inference from Illumina amplicon data). Non-singleton amplicon sequence variants (ASVs) were aligned with MAFFT and used for phylogeny via FastTree2. α-diversity (e.g., Chao1, Shannon, Simpson) and β-diversity metrics (e.g., weighted UniFrac, Jaccard distance) were estimated with the diversity plugin, standardizing to 20,000 sequences per sample. Lastly, taxonomy was designated to ASVs against the Greengenes 13.8 database (https://data.qiime2.org/2018.11/common/gg-13-8-99-515-806-nb-classifier.qza) via the q2-feature-classifier plugin.

Bioinformatics analysis

Data analysis was performed on an HP computing cluster using open-source tools in Linux. Biological data processing, statistical model formulation, and database creation were completed using these tools. QIIME2 and R packages (v3.2.0) were used for sequence data analyses. Α-diversity indices were computed using the ASV table in QIIME2. ASV-level ranked abundance curves were generated for sample comparison. Β-diversity analysis was done using UniFrac distance metrics, with results visualized via PCoA. PCA was carried out based on genus-level compositional profiles. ANOSIM was used to evaluate the significance of microbiota structure differentiation among groups. LEfSe was performed to identify differentially abundant taxa across groups.

Sample size estimation

This study was exploratory in nature; therefore, no formal sample size calculation was performed prior to recruitment. The sample size of 90 participants (30 per group) was selected based on feasibility and alignment with previous microbiome studies of similar design. Future studies should include power analysis and larger cohorts to validate the findings.

Statistical analysis

Data was processed using SPSS software (version 24.0) and graphical representations were created with GraphPad Prism software (version 7.0). The Shapiro-Wilk test was used to verify the normality of data distributions. Data was expressed as mean ± SD for normal distribution, or as median for non-normal distribution. Student’s t-test or one-way analysis of variance was used for normally distributed data, while the Wilcoxon’s rank-sum test was used for non-normally distributed data. Pearson’s χ2 test or Fisher’s exact test were used for rate comparisons. All tests were two-sided and significance level was set at P < 0.05.

Result

Participants characteristics

No statistically significant differences were found in maternal age, gestational age, or pre-pregnancy body mass index among healthy, HDP, and GDM pregnancies. Likewise, differences in adverse pregnancy outcomes, including placental abruption, fetal growth restriction, and fetal distress, were not statistically significant among the three groups. Gestational weight gain, birth weight, and gestational week were lower in HDP group compared to the normal group (all p < 0.05), while the cesarean section rate and blood pressure were higher in HDP group (Table 1).

Table 1.

Baseline demographic characteristics, clinical history, and adverse pregnancy outcomes of study participants

Factors	Healthy pregnant (n = 30)	HDP (n = 30)	GDM (n = 30)	P value
Demographic characteristics
Maternal age (year)	30.51 ± 4.18	30.54 ± 4.11	30.99 ± 4.69	0.89
BMI before pregnancy (kg/m2)	21.76 ± 3.54	23.56 ± 3.54	25.37 ± 2.47	< 0.001
Medical history
Family history of DM n (%)	0(0)	1(3.33)	5(16.67)	0.023
Family history of hypertension n (%)	0(0)	4(13.33)	3(10.00)	0.13
≥ 2 Gravidity time n (%)	12(40.00)	12(40.00)	15(50.00)	0.67
≥ 1History of term delivery n (%)	10(33.33)	11(36.67)	9(30.00)	0.86
≥ 1Spontaneous miscarriage n (%)	1(3.33)	4(13.33)	5(16.67)	0.23
Scarred uterus n(%)	7(23.33)	12(40.00)	14(46.67)	0.15
Maternal adverse outcomes
SBP	115.57 ± 19.28	141.44 ± 19.38	117.40 ± 17.63	< 0.001
DBP	73.52 ± 13.37	90.30 ± 13.52	73.54 ± 12.94	< 0.001
OGTT 0 h ≥ 5.1mmol/L (%)	0 (0)	0 (0)	30 (100)	/
OGTT 1 h ≥ 10 mmol/L (%)	0 (0)	0 (0)	16 (53.33)	/
OGTT 2 h ≥ 8.5 mmol/L (%)	0 (0)	0 (0)	18 (60.00)	/
Parturient at admission n (%)	28(93.33)	28(93.33)	27(90.00)	0.86
PROM n (%)	4(13.33)	7(23.33)	6(20.00)	0.60
Caesarean section n (%)	5(16.67)	17(56.67)	14(46.67)	0.004
Placental abruption n(%)	0(0)	0(0)	0(0)	1.00
Infants adverse outcomes
Stillbirth n(%)	0(0)	0(0)	0(0)	1.00
Premature delivery < 37wk n(%)	0(0)	10(33.33)	7(23.33)	0.003
Intrauterine distress n(%)	0(0)	0(0)	0(0)	1.00
IUGR n(%)	0(0)	3(10.00)	2(6.67)	0.23
SGA n(%)	0(0)	2(6.67)	0(0)	0.13
Low birth weight infant < 2500 g	1(3.33)	8(26.67)	0(0)	< 0.001
Birth weight (g)	3155.36 ± 615.25	2700.65 ± 631.62	3496.15 ± 740.21	< 0.001

BMI Body Mass Index, HDP Hypertensive disorders, GDM Gestational diabetes in pregnancy, SBP Systolic blood pressure, DBP Diastolic blood pressure, OGTT Oral Glucose Tolerance Test, PROM Premature rupture of membranes, IUGR Intrauterine growth retardation, SGA Small for Gestational Age

Changes in the composition of the gut microbiota

In this study, 90 fecal samples were collected for sequencing. At the phylum level, the relative abundance of the top 10 taxa indicated that the dominant bacterial phyla were Firmicutes, Actinobacteria, Bacteroides, Proteobacteria, and Verrucomicrobia. The findings revealed a significantly higher abundance of Actinobacteria in HDP women comparison to normal control (p < 0.05). Conversely, HDP group exhibited a lower abundance of Bacteroidetes (Fig. 2). In GDM group, the abundance of Bacteroidetes was significantly elevated compared to the normal control group (p < 0.05), with a lower abundance of Veillonellaceae (Fig. 2). At the family level, the abundance of Pasteurellaceae was increased in HDP group, with a low abundance for Bifidobacteriaceae. Furthermore, the abundance of Synergistes was much higher in both HDP and GDM group when comparing to normal control (Fig. 3).

Fig. 2.

Changes in the composition of the gut microbiota in the diseased and healthy control groups. A The abundance of Bacteroidetes was decreased, and the abundance of Actinobacteria was increased in HDP group compared to the control group. B The abundance of Bacteroidetes was higher in GDM group compared to the control group. HDP, hypertensive disorders of pregnancy. GDM, gestational diabetes mellitus

Fig. 3.

Analysis of microbial composition at the family level. Bifidobacteriaceae and Pasteurellaceae were significantly decreased and increased, respectively, in the HDP group compared to the normal control group (p < 0.05). Synergistaceae showed a significant increase in both the HDP and GDM groups compared to the normal control group (p < 0.05). HDP, hypertensive disorders in pregnancy; GDM, gestational diabetes mellitus

We employed PCoA to discern potential differences between the disease and control groups. However, the results indicated that HDP, GDM, and controls were not distinctly differentiated (Fig. 4). Additionally, we examined the alpha diversity of fecal microbes within these groups. Figure 5 reveals that, in comparison to the control group, the microbial alpha diversity was marginally lower in the HDP and GDM group, as indicated by the Chao 1 diversity index, Shannon diversity index, Observed species diversity index, and PD whole tree. However, the differences were not statistically significant except in the GDM group on chao1 and PD whole tree (Fig. 5).

Fig. 4.

The PCA Analysis. The results showed that the two groups cannot be completely distinguished from each other. PCA, principal component analysis

Fig. 5.

Analysis of microbial alpha diversity in disease and control groups. ns = no significant, *P < 0.05

Bacteria taxa differences between the HDP, GDM and the normal pregnancy

To further investigate which taxa can serve as the biomarker and have the ability to distinguish the HDP patients from the healthy controls, we used the LEfSe analysis to explore the different changes and relative richness of the gut microbiota. A total of 24 abundant taxa were different in the HDP group and normal group, which had a log LDA score > 2. The relative abundances of Bacteroidetes, Methanosphaera, Parabacteroides, Porphyromanadaceae, Rikenellaceae, Christensenellaceae, and Clostridium were lower in the HDP group than in the control group, while the relative abundances of Brevibacterium, Dermacoccus, Bifidobacterium, and Actinobacteria were higher in the HDP group. Similarly, the relative abundances of Dermacoccus, Prevotella, Gemellaceae, Leuconostocaceae, and Clostridium were lower in the GDM group than in the control group, while the relative abundances of Bifidobacterium, Actinobacteria, and Mogibacterium were higher in the GDM group (Fig. 6). Notably, the high abundances of Bifidobacterium and Actinobacteria in both GDM and HDP group illustrated that they may played essential roles in adverse progress of pregnant women.

Fig. 6.

Analysis through LEfSe. A Compared to the control group, the HDP group has a higher abundance of Actinobacteria and a lower abundance of Bacteroidia. B Compared to the control group, the GDM group has a higher abundance of Bifidobacteriaceae and Actinobacteria, and a lower abundance of Veillonellaceae and Roseburia. HDP, hypertensive disorders of pregnancy; GDM, gestational diabetes mellitus

Logistic regression model

Upon conducting a multiple logistic regression analysis with seven microbial variables, three significant predictors were identified at the α = 0.1 level: Bacteroidetes, Actinobacteria, and Synergistes. A logistic regression model was developed based on these taxa. Initial receiver operating characteristic (ROC) curve analysis showed that the area under the curve (AUC) for predicting HDP and GDM was 0.65 and 0.58, respectively, indicating low predictive accuracy when using microbiota data alone (Fig. 7).

Fig. 7.

The area under the curve of logistic regression model

However, when Body Mass Index (BMI) was included in the model alongside the three microbiota predictors, the AUC values improved to 0.77 for HDP (95% CI: 0.65–0.88; sensitivity: 82.3%, specificity: 66.7%) and 0.69 for GDM (95% CI: 0.56–0.82; sensitivity: 80.7%, specificity: 59.8%). These results suggest that while the microbiota-based model alone has limited predictive power, the addition of BMI enhances the diagnostic performance to a moderate level, particularly for HDP.

Discussion

As a diverse microbial community residing in the human intestine, the gut microbiota is abundant and varied in species, and its role in human disease development has been extensively studied [16]. Prior research has confirmed that changes in the abundance and diversity of the gut microbiota can contribute to the onset of metabolic disorders like diabetes and obesity [17, 18]. Importantly, HDP and GDM are prevalent complications during pregnancy. The gut microbiota is associated with the development of chronic diseases [19]. However, the underlying mechanisms linking the intestinal microbiota to HDP and GDM in pregnant women remain unclear. This study aims to explore the relationship between the intestinal microbiota of HDP and GDM in pregnant women and investigate potential mechanism involved.

Through 16S rRNA sequencing, this study revealed significantly difference in the diversity and composition of the gut microbiota for HDP and GDM group in before 20 weeks of pregnancy. Bacteroidetes phylum showed a highest abundance in GDM group when comparing to HDP group and normal control. Similarly, Gao et al. also found that Bacteroidetes had an increased abundance in GDM population, which was positively correlated with Hemoglobin A1C (HbA1C) [20]. Through the systematic review, Kunasegaran has summarized that the change of Bacteroidetes ratio could remarkedly affect the cell metabolism in pregnant women [21]. More importantly, Aatsinki et al. [22] and Wu et al. [23] had demonstrated a higher abundance of Bacteroidetes in middle and advanced GDM group, respectively. Thus, it can be assumed that Bacteroidetes could participate in gut mediated glucose metabolism throughout the entire pregnancy cycle. Similarly, our result reported that Actinobacteria displayed a higher in HDP group (p < 0.05). However, a previous meta-analysis including 19 articles reported that the relative abundance of Actinobacteria phylum in gut microbiota had no statistical difference between hypertension and healthy controls [24]. But Li has found that Actinobacteria phylum had a significant increase in hypertensive rats that correlated with host-microbiota inflammatory regulation in hypertension [25]. Besides, in Hsu’study, phylum Actinobacteria was significantly decreased in anti-hypertensive treatment for spontaneously hypertensive pregnant rats by using N-acetylcysteine, illustrating its important role in GDM [26]. There were still no relevant studies to confirm the role of Actinobacteria in GDM women. Nevertheless, Zheng et al. has suggested that increasing the abundance of Actinobacteria was reliable to alleviate gestational diabetes by probiotic supplements [27]. Hence, Actinobacteria phylum may serve as an underlying treated target for GDM women. Conversely, Synergistes family was found markedly higher in both GDM and HDP group in our results. Synergistes group severs as a special phylogenetic cluster of Gram-negative anaerobes, which was widely distributed in human mouth and gut [28]. Currently, it is lacked of enough evidence to suggest that Synergistes family in gut is related to adverse reactions in pregnant women. But some research reported that its imbalance may be an important factor affecting intestinal inflammation that dramatically associated with hypertension and diabetes mellitus [29]. Therefore, Synergistes family may be a new field for studying the mechanism of HDP and GDM. Furthermore, our results also suggested that Alpha and beta diversity comparisons indicated differences in gut microbiota richness among the HDP, GDM, and control groups. However, PCA analysis did not reveal a clear distinction between the groups, which may be limited by insufficient sample size in our study.

Several previous studies have examined the relationship between gestational diabetes and gut microbiota [30, 31]; however, our study is among the first to propose a predictive model based on gut microbiota collected prior to the 20th week of pregnancy, offering a novel approach to early screening. While the relationship between hypertension and intestinal microbiota has also been studied [32], fewer investigations have addressed this link in pregnant populations. Most related research focuses on placental [33] or vaginal microbiota [34]. Our prospective cohort study contributes by identifying early microbial signatures potentially associated with HDP and GDM before clinical onset, thus providing insight into early risk stratification. We developed a logistic regression model incorporating microbial predictors and plotted ROC curves to evaluate diagnostic performance. The area under the curve (AUC) improved from 0.65 (HDP) and 0.58 (GDM) using microbial variables alone, to 0.77 (95% CI: 0.66–0.88) for HDP and 0.69 (95% CI: 0.56–0.82) for GDM when BMI was included. While these values suggest moderate predictive utility, they indicate that intestinal microbiota—particularly in early pregnancy—may aid in risk prediction. Nevertheless, integration of traditional clinical indicators and metabolic biomarkers is essential to further improve predictive accuracy.

This research not only enhances our understanding of the intestinal microbiota’s functionality in pregnant women but also enables early prediction and intervention of HDP and GDM by regulating the intestinal microbiota. It has significant implications for preventing and reducing metabolic disorder-related mortality during pregnancy and alleviating healthcare burdens. Additionally, the disturbance in the gut microbiota can induct a substantial impact on the development of GDM and HDP. Supplementation with microbiota modulators can improve intestinal microbiota dysbiosis, providing novel targets and strategies for the prevention and treatment of GDM and HDP.

This study prospectively examined gut microbiota before 20 weeks of gestation in relation to later development of GDM and HDP, providing early insight into potential microbial markers. The use of strict exclusion criteria and adherence to gestational weight gain guidelines helped reduce confounding. However, the relatively small sample size may limit statistical power, and PCoA did not reveal distinct group separation. Dietary data were not standardized, which may introduce confounding, and the predictive models lack external validation. Future studies should address these limitations in larger, well-controlled cohorts.

Conclusion

In conclusion, this study represents the first attempt to utilize early gut microbiota for predicting and providing early warnings of gestational metabolic disorders. Despite the less-than-ideal results, potentially due to the small sample size, this research offers a novel strategy for future prediction of gestational metabolic disorders and holds considerable significance.

Abbreviations

GDM

Gestational diabetes mellitus

HDP

Hypertensive disorders in pregnancy

SBP

Systolic blood pressure

DBP

Diastolic blood pressure

OGTT

Oral glucose tolerance test

ASVs

Amplicon sequence variants

HbA1C

Hemoglobin A1C

PCA

Principal component analysis

AUC

The area under the curve

Authors’ contributions

ZHZ and SSZ designed this study, performed the research, analyzed the data and wrote the paper. XYO helped to search and collect the data. LJD and YCG worked on design and supervision of review, funding acquisition, and project administration. All the authors have read and approved the final manuscript.

Funding

This study was supported by Youth Innovation Research Project Plan of Sichuan Medical Association (No. Q23062), the Applied Basic Research Foundation of Sichuan Provincial Science and Technology Department (No. 2021JY0240), Sichuan Science and Technology Program (No.2022YFS0623 and No.2023ZYD0287), Luzhou Municipal People’s Government and Southwest Medical University (No. 2021LZXNYD-D01), Health Commission of Sichuan Province (No.21ZD006), Medical Research Project of Sichuan Medical Association (No.S21038), and Southwest Medical University (No. 2023ZYQJ01).

Data availability

The raw sequencing dataset analyzed in this study has been deposited in the European Bioinformatics Institute (EBI) database under the accession code PRJEB31743 (https://www.ebi.ac.uk/ena/data/view/PRJEB31743). The OTU and taxonomic composition data, and the statistical scripts are available from the corresponding authors upon reasonable request.

Declarations

Ethics approval and consent to participate

The ethics committee of Guangzhou Women and Children’s Medical Center approved all aspects of this study (ID: 2018030306) and informed consent to participate was obtained from all of the participants.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.