The ultrasound screening is performed by the direct visualization of the umbilical cord or by tracking the umbilical arteries around the fetal bladder with color Doppler technology (10). The diagnosis of UAT is established when an initial ultrasound examination in the first trimester of pregnancy indicates normal umbilical artery flow, but later scans reveal the presence of a single umbilical artery.

Upon staining with Hematoxylin–Eosin, sections of the umbilical cord revealed the presence of thrombosis (specifically, fibrinous, mixed, or red thrombus) within one of the umbilical arteries. This thrombosis exhibited features that ranged from total or partial necrosis of the artery wall to cases where no evident necrosis was observed (11).

The study population was assigned into two distinct groups: the experimental group and the expectant group, based on the treatment regimen they received. Participants in the experimental group received LMWH (nadroparin calcium 4,100 U daily or enoxaparin sodium 4,000 U daily) for anti-coagulation therapy, underwent rigorous ultrasound surveillance, non-stress testing beyond 28 weeks of gestation, and were instructed to closely monitor fetal movements. In contrast, participants in the expectant group received only standard prenatal care.

Data pertaining to maternal age, gravidity, parity, gestational weeks at diagnosis, gestational weeks at delivery, and neonatal outcomes such as birth weight, Apgar scores, cesarean delivery, fetal distress, neonatal morbidity, and newborn intensive care unit (NICU) admission were collected and analyzed.

Statistical analysis was carried out with SPSS 25.0 for Microsoft Windows (IBM Corp., Armonk, NY, USA). Comparisons of continuous variables between groups were performed using Student’s t-test (for data that were normally distributed) or Mann–Whitney U-test (for data that exhibited non-normal distribution), while comparisons of categorical variables were conducted with the χ² test. Multi-factor ANOVA was used to investigate the effects of independent variables on birth weight. Multivariable logistic regression models were utilized to explore the association between the independent variables and birth weight. A p-value of < 0.05 was considered statistically significant.

During the study period, a total of 182,942 deliveries were recorded. Among these, 65 pregnancies were diagnosed with UAT. However, eight cases were subsequently excluded due to fetal death at presentation. Finally, a total of 57 participants with UAT were included in the analysis, with 29 participants assigned to the experiment group and the remaining 28 designated to the expectant group. Details are presented in Figure 1.

The results of the comparison pertaining to baseline clinical characteristics between the experiment group and the expectant group are presented in Table 1. Consequently, no significant differences were noted in terms of maternal age (30.14 ± 4.23 vs. 30.79 ± 3.92, p = 0.552), BMI (26.75 ± 4.99 vs. 26.05 ± 3.09, p = 0.527), gravidity (1.97 ± 1.38 vs. 2.07 ± 1.33, p = 0.769), primigravida (48.28% vs. 46.43%, p = 1.000), parity (0.41 ± 0.68 vs. 0.43 ± 0.57, p = 0.930), primipara (31.03% vs. 39.29%, p = 0.585), and gestational age at diagnosis (29.90 ± 4.62 vs. 31.00 ± 4.31, p = 0.497).

The results of the comparison of clinical outcomes between the experiment group and the expectant group are presented in Table 2. The birth weight in the experiment group was significantly lower compared to the expectant group (1874.46 ± 717.83 vs. 2434.40 ± 770.20, p = 0.008), while the proportion of births before 34 weeks was significantly higher in the experiment group than the expectant group (82.75% vs. 42.24%, p = 0.001), indicating a higher prevalence of early preterm delivery (<34 weeks of gestation) and lower birth weight in the experiment group. Gestational age at birth was notably lower in the experiment group as compared to the expectant group, although the difference did not reach statistical significance. No significant differences were noted in terms of Apgar scores, cesarean delivery, fetal distress, neonatal morbidity, and NICU stay.

The results of the multi-factor ANOVA are summarized in Table 3. The ANOVA revealed statistically significant anti-coagulation therapy (LMWH for participants in the experiment group) (F = 4.479, p = 0.039) and gestational age at birth (F = 179.110, p = 0.000) on birth weight.

The results in Table 4 indicated a significant negative association of anti-coagulation therapy (β = −0.164, p = 0.014) and a strong positive correlation of gestational age (β = 0.846, p < 0.001) with birth weight. The final linear regression equation can be expressed as: birth weight = −3314.782–256.106 × anti-coagulation therapy +161.858 × gestational age at birth (weeks).

The key findings of our study were: (1) expectant therapy had a significant advantage over experiment therapy in pregnancies with UAT, as the administration of LMWH resulted in a significant decrease in birth weight and a substantial increase in the incidence of preterm birth (<34 weeks); (2) the relationship between these variables can be formulated as: birth weight = −3314.782–256.106 × anti-coagulation therapy +161.858 × gestational age at birth. The key findings are incorporated and presented in Figure 2.

UAT may occur after placenta thrombotic vasculopathy (12), umbilical cord abnormalities (3, 13, 14), and underlying maternal diseases (2, 15). At present, no consensus has been reached on the treatment strategy for UAT. In clinical practice, upon the occurrence of a fetal umbilical cord embolism, one would instinctively consider the implementation of anti-coagulation therapy to arrest the progression of thrombus emboli, thereby preventing complete occlusion of the umbilical vessels and reducing the subsequent risk of fetal death. For instance, Wang et al. (6) reviewed 10 cases of pregnancies with UAT. Notably, all participants in the study received treatment with LMWH, and there was no control group of expectant mothers. Based on their findings, the researchers concluded that the early administration of LMWHs may enhance pregnancy outcomes. However, there have been arguments raised against the use of anticoagulation treatment. Wei et al. reported a case series revealing that the expectant management of UAT had comparable fetal outcomes to those observed in patients who received anti-coagulation management (3). Han et al. demonstrated that expectant treatment of patients with UAT had apparent positive effects for extending gestational age, which was supported by another study conducted by Dindinger et al. (5, 9). Our study found that expectant management had significant advantages compared to anti-coagulation therapy, as the latter, when combined with frequent ultrasound surveillance, could induce anxiety in both patients and clinicians, resulting in unnecessary early medical interventions to deliver the fetus.

Our study has several strengths. First, it was designed as a retrospective study with two arms (expectant vs. experiment), whereas the previous studies (2–4, 6, 7) were primarily case reports. Second, we used multi-factor ANOVA to analyze the associations between the expectant group and the experiment group, leveraging its advantages in accounting for the potential confounding effects of multiple variables. Third, due to the rigorous design and comprehensive analysis, our results were more aligned with real-world clinical logic and practical experiences, indicating a greater degree of applicability and reliability. However, it is important to acknowledge that the retrospective design of our study, coupled with the relatively small sample size, may result in incomplete data and an increased risk of bias.