First-trimester vaginal microecology and spontaneous abortion risk: retrospective cohort
First-trimester vaginal dysbiosis — defined by abnormal bacterial density, Nugent score ≥7, and Lactobacillus depletion — independently associates with significantly increased spontaneous abortion risk.
| Population | First-trimester pregnant women attending a tertiary obstetric hospital in China (2023–2025), mean age 30.5 years |
|---|---|
| Exposure | Vaginal microecology assessment (bacterial density and diversity, Nugent score, Lactobacillus grade, pH, vaginal cleanliness, leukocyte esterase, sialidase, catalase, Donders score) during the first trimester |
| Comparator | Women without abortion (delivery group, n=626) versus women with spontaneous abortion (n=160) |
| Outcome | Spontaneous abortion — abnormal bacterial density; Spontaneous abortion — Lactobacillus depletion (LBG IIb-III); Spontaneous abortion — bacterial vaginosis (Nugent ≥7); Spontaneous abortion — abnormal vaginal pH; Spontaneous abortion — abnormal vaginal cleanliness; Mean Lactobacillus proportion (% of total bacterial morphotypes); Spontaneous abortion — aerobic vaginitis (AV, Donders ≥3) |
Summary of findings
| Outcome | Effect | 95% CI | Certainty | Clinical relevance | Notes |
|---|---|---|---|---|---|
| Spontaneous abortion — abnormal bacterial density | OR 3.75, 95% CI 2.21-6.33 | — | Low | — | 1 studies |
| Spontaneous abortion — Lactobacillus depletion (LBG IIb-III) | OR 3.44, 95% CI 2.37-4.99 | — | Low | — | 1 studies |
| Spontaneous abortion — bacterial vaginosis (Nugent ≥7) | OR 2.77, 95% CI 1.56-4.92 | — | Low | — | 1 studies |
| Spontaneous abortion — abnormal vaginal pH | OR 1.74, 95% CI 1.21-2.49 | — | Low | — | 1 studies |
| Spontaneous abortion — abnormal vaginal cleanliness | OR 1.91, 95% CI 1.33-2.74 | — | Low | — | 1 studies |
| Mean Lactobacillus proportion (% of total bacterial morphotypes) | MD -19.1 percentage points (42.5% vs 61.6%, p<0.001, in the CI reported) | — | Low | — | 1 studies |
| Spontaneous abortion — aerobic vaginitis (AV, Donders ≥3) | OR not fully reported in available text; p-value not extractable | — | Low | — | 1 studies |
Context
Approximately 50% of spontaneous abortions remain unexplained. Reproductive tract infections account for up to 15% of first-trimester losses. Clinical data linking conventional vaginal microecology parameters to abortion risk in Asian populations are scarce.
What the study showed
Abnormal bacterial density associated with abortion (OR=3.75; 95% CI 2.21–6.33; p<0.001). Lactobacillus depletion (LBG grade IIb-III) showed OR=3.44 (95% CI 2.37–4.99; p<0.001). Nugent score ≥7 (bacterial vaginosis) associated with abortion (OR=2.77; 95% CI 1.56–4.92; p<0.001). Mean Lactobacillus proportion was 42.5±35.85% in the abortion group versus 61.6±26.72% in the non-abortion group (p<0.001). All models adjusted for maternal age.
How it was done
Single-centre retrospective cohort; 798 first-trimester women screened for vaginal microecology; after exclusions (n=12), 786 participants analysed (160 abortions, 626 deliveries). Data collected at a tertiary obstetric hospital in Fujian, China, 2023–2025. Multivariable analysis adjusted for maternal age; chi-square or Fisher's exact test for categorical variables.
Effect magnitude
Largest effects: abnormal bacterial density (OR=3.75; 95% CI 2.21–6.33) and Lactobacillus depletion (OR=3.44; 95% CI 2.37–4.99), both p<0.001 after maternal age adjustment.
Risk of bias
Retrospective design precludes causal inference. Single-centre study with predominantly Han population (96.9%) limits generalisability. No genomic sequencing — species-level Lactobacillus identification impossible. Adjustment only for maternal age; key confounders (BMI, parity, prior antibiotic use, genetic cause of abortion) uncontrolled. Risk-of-bias tool not declared by authors (ROBINS-I would apply). Genetic aetiology of abortions not systematically excluded.
What this study does NOT prove
This study does not prove causality between vaginal dysbiosis and spontaneous abortion — association is not causation. Findings are not generalisable to other ethnic groups, different socioeconomic contexts, or healthcare systems outside China.
In clinical practice
Conventional vaginal microecology screening (Nugent score, pH, Lactobacillus grade) at first-trimester prenatal visits can identify women with dysbiosis at elevated abortion risk. Clinicians should consider treating documented BV (Nugent ≥7) in the first trimester per current guidelines. The statistical association does not establish causality; therapeutic decisions must integrate full clinical context.
Limitations
Retrospective design precludes causal inference. Single-centre study with predominantly Han population (96.9%) limits generalisability. No genomic sequencing — species-level Lactobacillus identification impossible. Adjustment only for maternal age; key confounders (BMI, parity, prior antibiotic use, genetic cause of abortion) uncontrolled. Risk-of-bias tool not declared by authors (ROBINS-I would apply). Genetic aetiology of abortions not systematically excluded.
What is still missing
Prospective studies with metagenomic sequencing and systematic control for genetic causes of abortion are needed to establish causality and identify protective species. Randomised controlled trials evaluating first-trimester dysbiosis intervention on abortion outcomes are lacking.
Technical appendix
Version history
- 1.0 · 2026-11-07 — Auto-generated under Evidence Standard v1.0