The association of serum total bile acid with new-onset hypertension during pregnancy

As shown in Fig. 1, we collected data from women who underwent routine gestational care from 2014 to 2020 at the Third Affiliated Hospital of Guangzhou Medical University. There were 9,677 singleton deliveries with TBA tests, and patients with chronic hypertension, hepatitis B or C, cholecystopathy, inflammation, a diagnosis of ICP, serum TBA levels ≥ 10 µmol/L, other terminal diseases such as phase iv cancer, end-stage renal disease, and AST or ALT levels over 100 U/L were excluded. A total of 1071 women were included in the non-HDP group, and 1510 women were included in the HDP group and further divided into PE (n = 1030) and GH (n = 480) subgroups.

Maternal clinical parameters were collected from medical record, including maternal age, gestational age at delivery (GA), body mass index (BMI), systolic blood pressure (SBP), diastolic blood pressure (DBP), whether receive in-vitro fertilization (IVF), nulliparity, pregestational diabetes infant birth weight and body length. SBP and DBP data in HDP group were collected for the first time measured after diagnosis, and in the non-HDP group, SBP and DBP were collected for the first time measured after hospitalized. Maternal fasting blood samples were collected by first-time examination after hospitalization, after an overnight fast of at least 8 h at a mean gestational age of 20 to 34 weeks. The biomarkers included TBA, ALB, TC, TG, AST, ALT, T bilirubin, and 24-h proteinuria. TBA, TC, TG, AST, ALT were assessed by Roche cobas^@8000 modular analyzer series (Mannheim, BW, Germany), and T bilirubin, ALB and 24-h proteinuria Roche cobas^@ C501(Mannheim, BW, Germany).

The criteria suggested by the American College of Obstetricians and Gynecologists (ACOG) for the diagnosis of pregnancy-associated hypertension were adopted. PE was defined as systolic blood pressure ≥ 140 mmHg or diastolic blood pressure ≥ 90 mmHg over at least two intervals of 6 h after 20 weeks of gestation with previously normal blood pressure and new-onset proteinuria (> 0.3 g/24-h or at least > 1 + on protein dipstick when urine could not be collected for 24 h). And liver function impaired, thrombocytopenia, severe persistent right upper quadrant or epigastric pain and not accounted for by alternative diagnoses, renal insufficiency, pulmonary edema, new onset headache unresponsive to acetaminophen and visual disturbances should diagnosed as PE for women with gestational hypertension in the absence of proteinuria [3]. Severe PE was defined as systolic blood pressure ≥ 160 mmHg or diastolic blood pressure ≥ 110 mmHg and new-onset headache or Visual disturbances. Early-onset PE (EOPE) was defined as development PE before 34 weeks of gestation and late-onset PE (LOPE) was defined as development PE after 34 weeks of gestation [18].

The primary pregnancy outcomes in our study were low birth weight and fetal growth restriction (FGR). FGR was diagnosed following the guidelines for FGR [19], as indicated by ultrasound-estimated fetal weight or circumference of the abdomen smaller than 10% of the birth population. Low birth weight (LBW) was defined as birth weight less than 2500 g.

Data were presented as the mean ± standard deviation or median (25–75 percentile). According to the normality of the variable distributions, measurement data were expressed as the means ± standard deviations, except for GA, SBP, DBP, TBA, AST, ALT, 24 h proteinuria and birth length. An independent-samples t-test was performed for age, BMI, ALB, TC, TG and birth weight. Count data were expressed as proportions (%), and the Mann–Whitney U test was performed for GA, SBP, DBP, TBA, AST, ALT, 24 h proteinuria and birth length. Spearman correlation analysis was used to evaluate the correlation between maternal serum TBA levels and birth weight and between AST/ALT and TBA levels. Multiple logistic regression analysis was performed to examine the influence of the following variables: maternal age, prepregnancy BMI, AST, ALT, T bilirubin, TC, TG, ALB, pregestational diabetes, nulliparity and IVF. The odds ratios with 95% CIs were calculated.

This study had the following findings: 1) Maternal serum TBA was significantly and positively associated with HDP and severity of PE; 2) The TBA level was negatively related to birth weight and FGR. Taken together, these findings indicate that maternal serum TBA is a potential prognostic biomarker of PE.

Intrahepatic cholestasis of pregnancy (ICP) was determined by pruritus occurring during pregnancy and serum TBA above 10 µmol/L, and excluding for elevated serum aminotransferase and/or serum bile acids related to other causes of liver test abnormalities [24, 25]. ICP has been reported to be related to maternal and fetal diseases during pregnancy, such as PE, gestational diabetes mellitus (GDM) [26], placental abnormalities [27] and stillbirth [28]. However, the relationship between HDP and TBA levels in the normal range remains unclear.

In our study, we demonstrate that maternal serum TBA levels in the normal range were significantly higher in the HDP group than those in the non-HDP group and could be an independent risk factor for HDP. In addition, the level of serum TBA is significantly and positively associated with the severity of PE. Although, total bile acid was not significant change during pregnancy, especially second and third trimesters [29, 30]. High levels of bile acid have been reported to cause placental structural damage due to vasoconstriction of the chorionic plate veins [31] and affect the placental antioxidant system, which leads to oxidative stress and increases placental syncytial knots [32]. An increase in the number of placental syncytial knots was reported in PE [27]. Furthermore, although the pathophysiology of PE is not fully understood, increased maternal endoglin and soluble-fms-like tyrosine kinase-1 (s-Flt1) levels are known to be major contributors to PE. Oxidative stress was reported to increase the expression of s-Flt1 by trophoblast cells of the placenta [33]; thus, increased serum bile acid-induced oxidative stress might upregulate the expression of s-Flt1 and could therefore explain our observation. Moreover, we found maternal TBA level was positively related to SBP, DBP and proteinuria. There were studies that reported that bile acid had a vasodilation effect [34, 35]. These studies are more focused on the secondary bile acid, especially Ursodeoxycholic acid (UDCA) and lithocholic acid (LCA), which were predominant in liver cirrhosis. However, other bile acid derivates, such as cholic acid, have less effect on vasodilation, and marinobufagenin (MGB) elevated blood pressure by inhibiting vascular NA/K-ATPase [36, 37].

Fetal complications, especially LBW and FGR, have been reported in PE and ICP pregnancies. A recent study in China reported increased TBA levels with a decrease in birth weight [38]. Song and coworkers also found that serum TBA levels in all ranges were associated with the risk of FGR and that HDP had an additive effect on the association [24]. Consistent with these studies, our data show that the maternal TBA level was negatively associated with birth weight and positively associated with the risk of LBW and FGR. Besides, previous reports more focused on the relationship of LBW and FGR with all range TBA or even TBA levels in ICP patients [24, 38]. Based on our study, even TBA level in the normal range, the elevation is also positively related to the risk of LBW and FGR. Bile acid accumulation may affect placental vascular remodeling and thus cause insufficient placental perfusion, leading to fatal nutritional deficiency and FGR [39, 40]. Furthermore, toxic bile acid disrupts the expression of placental angiogenic and antiangiogenic factors, including s-Flt1 and endoglin, which have been reported to contribute to PE and FGR [41]. The oxidative stress caused by bile acid accumulation increases the expression of proinflammatory cytokines and chemokines, such as TNF-α, C-reactive protein and IL-8, in the placenta and maternal peripheral organs [42], which will activate the NF-κB pathway. Recent studies have revealed that placentas from FGR infants show strong Nf-κB p65 immunoreactivity [43], and animal experiments have also found maternal inflammation and oxidative stress in FGR [44]. Further studies are required to elucidate the pathological mechanism by which excessive bile acid causes FGR, especially in PE patients.

To our knowledge, this is the first study to investigate the relationship between HDP and maternal serum TBA levels in the normal range. In addition, we included hospitalized patients, which ensured the integrity of the information and allowed us to study the relationship between HDP and TBA. However, the present study has some limitations. First, this study is a single-center study. Second, because this study is a cross-sectional observational study, causality cannot be determined. Third, although blood sample were collected after overnight fasting, we do not measure hormone level, dietary habit and emotional status that are factors that can influence the TBA level. Last, bile acid has many secondary forms; in this study, we discuss only the total bile acid in serum.