Dysregulation of bile acids increases the risk for preterm birth in pregnant women

The study addresses whether dysregulated bile acids are an etiologic factor for preterm birth (PTB). PTB, defined as delivery before 37 weeks of gestation, is the leading cause of infant mortality and is associated with lifelong morbidities. Despite extensive efforts, mechanisms of PTB remain poorly understood and current preventive interventions show inconsistent benefit. Bile acids, synthesized in the liver and regulated by the enterohepatic circulation via farnesoid X receptor (FXR) signaling, are increasingly recognized as signaling molecules. Clinical observations in intrahepatic cholestasis of pregnancy (ICP) link elevated bile acids with increased PTB risk, but a direct causal relationship has not been established, nor is it known whether liver injury per se or elevated bile acids drive PTB. This study combines a large hospital-based cohort of pregnant women with mouse models to test the hypothesis that elevated circulating bile acids increase PTB risk and that restoring bile acid homeostasis can prevent or delay PTB.

Prior work has shown associations between elevated bile acids and adverse pregnancy outcomes, particularly in ICP, where increased serum bile acids correlate with PTB risk. Bile acids function as signaling molecules affecting diverse metabolic pathways through receptors including FXR. Advanced maternal age and high body mass index (BMI) have been linked epidemiologically to increased PTB risk, and non-pregnant studies report higher serum bile acids in obesity. However, an etiologic link between bile acids and PTB, and the role of liver injury versus bile acid elevation, had not been established before this work.

Human cohort: A hospital-based cohort of 36,755 pregnant women (predominantly ethnic Chinese, 99.6%) was analyzed. Participant characteristics included maternal age and pre-pregnancy BMI; parity and histories of PTB or stillbirth were recorded. Serum total bile acids (sTBAs) and liver function markers (AST, ALT, total bilirubin (TBiL), GGT) were measured at routine antenatal visits. PTB was defined as delivery before 37 weeks and categorized as spontaneous PTB (sPTB) or iatrogenic/medically indicated PTB (iPTB). Associations between sTBA (and liver markers) and PTB rates were analyzed overall and within strata of age and BMI. Subgroups with liver disorders, including nonalcoholic fatty liver disease (NAFLD) and ICP, were evaluated. Statistical analyses included Student’s t-test, one-way ANOVA with Tukey post hoc, chi-squared tests for proportions, Pearson correlations, and Poisson regression to estimate relative risk; p < 0.05 was considered significant.

Animal models: Multiple mouse models were used to probe causality.

• Estrogen-induced cholestasis model: Pregnant C57 mice were treated with 17α-ethinyl estradiol (EE2) beginning around gestation day (GD) 16.3. Outcomes included gestation length, PTB rate, live birth rate, sTBA, and AST.
• Hepatotoxic injury model: Pregnant mice received carbon tetrachloride (CCl4) (doses such as 1.0, 1.5, 2.0 ml/kg) around GD16.3 to induce liver injury and bile acid dysregulation; gestation length, PTB, live birth rates, sTBA, and AST were measured.
• Bile acid feeding model: Pregnant mice were fed diets containing cholic acid (CA) at 0.5% or 1% starting at approximately GD15.3 to elevate sTBA without overt liver injury; gestation, PTB, live birth rates, sTBA, and AST were assessed.
• Mechanism tests without overt liver injury: Non-pregnant mice were given 0.3% CA or an FXR antagonist (DY268) to elevate sTBA without increasing AST; pregnant mice received combined low-dose CA (0.3%) plus DY268 to elevate sTBA further and assess PTB without histological liver injury.
• FXR activation rescue: Pregnant mice treated with CCl4 were co-treated with the FXR agonist GW4064 to restore bile acid homeostasis; gestation length, PTB rates, live birth rates, sTBA, AST, and hepatic expression of bile acid homeostasis genes (Bsep, Cyp7a1, Shp, Fxr by qPCR) were measured. Histopathology assessed liver injury where indicated. Animal care and experimental approvals were as described; statistical approaches mirrored human analyses with appropriate tests.

Human cohort (n=36,755):

• PTB occurred in 2,875 women; 33,880 had full-term birth (FTB). Mean sTBA was higher in PTB versus FTB (5.69 ± 12.45 μM vs 4.68 ± 5.31 μM; p < 2.4×10^-18).
• As sTBA increased, PTB rates rose markedly. Compared with <10 μM, increasing sTBA categories showed higher PTB rates: approximately 2.76% (<10 μM) rising to 9.26% (10–39.99 μM; RR 1.21, 95% CI 1.05–1.39; p=0.0055), 29.6% (40–99.99 μM; RR 3.87, 95% CI 2.86–5.23; p=5.3×10^-14), and 50% (≥100 μM; RR 6.57, 95% CI 3.53–12.22; p=9.6×10^-14). Correlations indicated sTBA tracked with PTB rates.
• sPTB and iPTB subtypes both showed higher sTBA versus FTB; risks for sPTB and iPTB increased with higher sTBA. For sPTB, rates rose from 3.5% (<10 μM) to 30.3% (>409.99 μM; RR 29.59; p=0.001) and up to 80% at ≥2100 μM (RR 6.55; p=0.005). iPTB increased from 4.2% to 6.2% (10–39.99 μM; RR 1.51; p=7.6×10^-19), 12.9% (40–99.99 μM; RR 4.71; p=5.7×10^-21), and 30% (≥2100 μM; RR 7.37; p=2.7×10^-5).
• Liver injury markers associated with PTB: AST higher in PTB than FTB (20.35 ± 28.54 vs 16.36 ± 10.74 U/L). PTB rates increased with AST categories: 8.8% (0–40 U/L) to 9.6% (41–80 U/L; p=2.5×10^-5) and 27.7% (≥200 U/L; p=2.2×10^-17). Very high TBiL (>34.3 μM) and elevated GGT were also linked to increased PTB rates.
• Maternal age: Youngest (18–24) and advanced age groups (30–34, 35–39, 40–50) had higher PTB rates versus 25–29 years (6.8%), with sTBA trending upward with age and correlating with PTB (r≈0.98; p≈0.003).
• BMI: Compared with normal BMI (18.5–24.9; PTB ~7.1%), overweight (25–29.9) and obese (≥30) had higher PTB rates (~11.9% and ~14.6%, respectively), with sTBA trending higher across BMI groups.
• Liver disease subgroups: PTB rates were higher in NAFLD (~13.96%) and ICP (~23.17%) than overall (~7.82%). sTBA and AST were elevated in NAFLD and ICP; within these groups, PTB cases had higher sTBA than FTB.

Mouse models demonstrating causality:

• EE2-induced cholestasis: EE2 reduced gestation days (from ~19.9 ± 0.2 to 18.76 ± 1.5; p=0.013), induced PTB (58.3% in treated), markedly reduced live birth rate (mean 34.4% vs control; p=0.001). sTBA increased (7.49 ± 2.01 to 31.21 ± 3.12 μM; p=3.5×10^-14) and negatively correlated with gestation days and live birth rates; AST rose substantially.
• CCl4-induced injury: Dose-dependent PTB induction; 1.5–2.0 ml/kg yielded 100% PTB; live birth rate decreased (~26.3% at 1.5 ml/kg; p=0.0015). sTBA elevations correlated negatively with gestation days and live birth.
• Cholic acid feeding: CA induced PTB dose-dependently: 0.5% CA → ~50% PTB, 1% CA → 100% PTB; gestation days shortened. sTBA rose sharply (to ~174 μM with 0.5% and ~250 μM with 1% CA). AST changes with CA were minimal and within normal ranges, indicating minimal liver injury while PTB occurred, supporting bile acids as the driver.
• Elevated bile acids without overt injury: 0.3% CA or FXR antagonist DY268 in non-pregnant mice increased sTBA without raising AST. In pregnant mice, combined low-dose CA + DY268 induced PTB with elevated sTBA, normal AST, and no histologic liver injury.
• FXR agonist rescue: GW4064 co-treatment with CCl4 increased gestation days, reduced PTB from 100% to 38.5% (p=0.04), and improved live birth from 25.6% to 91.4% (p=0.0025). sTBA decreased (38.3 ± 4.1 to 28.8 ± 5.5 μM; p=2.4×10^-6) and AST decreased (122.1 ± 22.8 to 43 ± 15.3 U/L; p=3.7×10^-5). Hepatic gene expression changes (upregulated Bsep; downregulated Cyp7a1) indicated restored bile acid homeostasis.

The human cohort demonstrates that higher circulating bile acids are associated with increased PTB risk across maternal ages, BMI categories, and in the presence or absence of liver disease etiologies (NAFLD, ICP). Mouse models establish causality: elevating bile acids through cholestasis (EE2), hepatotoxic challenge (CCl4), direct bile acid supplementation (CA), or FXR antagonism precipitates PTB, while FXR activation mitigates risk by restoring bile acid homeostasis. Minimal AST elevations in the CA model, together with normal histology in the CA+DY268 model, indicate that elevated bile acids per se, rather than liver injury alone, can trigger PTB. These findings support an etiologic role for bile acids in PTB and suggest that sTBA may be a valuable biomarker to identify at-risk pregnancies. Therapeutically, targeting bile acid pathways (e.g., FXR agonism) could prevent or delay PTB and improve neonatal survival, particularly in women with dysregulated bile acids or liver disorders.

This study links dysregulated bile acids to the risk and causation of preterm birth. In a large human cohort, sTBA levels correlate with PTB rates, independent of maternal age, BMI, and liver disease etiology. Mouse experiments show that elevating bile acids induces PTB, whereas restoring bile acid homeostasis via FXR activation reduces PTB and markedly improves neonatal outcomes. These results establish bile acids as an etiologic factor in PTB and identify bile acid homeostasis as a tractable therapeutic target. Future work should elucidate the molecular pathways by which bile acids induce PTB, identify modifiers of individual susceptibility (given that some women with elevated bile acids did not deliver preterm), optimize FXR-targeted or other bile acid–modulating therapies for pregnancy, and evaluate predictive utility of sTBA in diverse populations.

The human data are observational from a single hospital-based cohort predominantly comprising ethnic Chinese women, which may limit generalizability. Some reported subgroup and figure values show variability, and correlation strengths vary across analyses. Although animal models support causality, translating dosing and exposures (e.g., EE2, CA, CCl4) to human pregnancy requires caution. Additionally, not all women with elevated bile acids experienced PTB, indicating the presence of other modifying factors and signaling pathway differences that remain to be identified.