This study was a retrospective cohort analysis involving 269 primiparous women who delivered macrosomic infants at Haidian Maternal and Child Health Hospital between January 2022 and February 2024. The inclusion criteria were: ① full-term, singleton pregnancy in primiparous women with cephalic presentation and normal pelvis; ② birth weight $\ge$4000 g; ③ complete medical records.

The exclusion criteria were: women with placenta previa, vasa previa, or a history of uterine surgery, all of which are indications for cesarean section.

The study was approved by the Ethics Committee of Haidian Maternal and Child Health Hospital (Approval No. 2025LW-14, valid from January 2022 to December 2024).

Data on maternal and neonatal outcomes were gathered, focusing on factors such as maternal age, pre-pregnancy body mass index (BMI), gestational weight gain, gestational hyperglycemia, gestational age at delivery, trial of labor, delivery method, reasons for cesarean section, intrapartum fever, shoulder dystocia, fetal stress, postpartum hemorrhage, neonatal sex, birth weight, and neonatal asphyxia. Among the participants, 160 had normal blood glucose levels, while 109 experienced hyperglycemia. Two participants with poor glycemic control were excluded from the analysis, resulting in a final count of 107 participants with well-controlled blood glucose levels in the hyperglycemic group and 160 participants in the normal blood glucose group. A comparison was made between the 2 groups regarding maternal demographic characteristics, delivery method, reasons for cesarean section (macrosomia, fetal stress, cephalopelvic disproportion) and delivery complications (shoulder dystocia, neonatal asphyxia).

Postpartum hemorrhage and gestational hyperglycemia were defined according to published guidelines [8, 9]. The diagnosis of gestational hyperglycemia was consistent with International Association of Diabetes and Pregnancy Study Groups (IADPSG) criteria. The diagnostic criteria for neonatal asphyxia are based on the Guidelines on Basic Newborn Resuscitation [10]. Intrapartum fever refers to maternal temperature $\ge$38 °C during labor, which may indicate complications such as intra-amniotic infection or epidural-related maternal fever [11].

Our hospital’s Ultrasound Department utilizes the Voluson E10 ultrasound equipment (GE Healthcare, Chicago, IL, USA) to assess fetal weight. Instead of relying on a single fixed formula, it automatically matches internationally recognized standardized formulas (such as the Hadlock series formulas, Warsofs formula, etc.) that have been validated through extensive clinical data, based on the fetal biological parameters measured by ultrasound (such as biparietal diameter, head circumference, abdominal circumference, and femur length). The selection of these formulas is usually determined by the fetal growth assessment system preset by the equipment, with the core goal of improving the accuracy of weight estimation through multi-parameter joint calculation. For example, Hadlock II formula may be preferred in early second trimester, while Warsofs formula may be preferred in late pregnancy.

We assumed the cesarean section rate in the hyperglycemia group was 57% and the cesarean section rate in the normoglycemia group being 40%, Significance level ($\alpha$): Typically set at 0.05 for a two-tailed test (corresponding to a critical value Z_α/2 = 1.96).

Power (1-$\beta$): It is recommended to be $\ge$80% ($\beta$ = 0.2, corresponding to Z$\beta$ = 0.84).

Two group proportions: p1 = 57% = 0.57 (Group A), p2 = 40% = 0.40 (Group B).

Rate difference: $\delta$ = p1 – p2 = 0.17 (17%).

The sample size calculation for comparing the difference in proportions between 2 groups is typically based on the normal approximation method, and its general formula is:

n = (Z_α/2 + Z$\beta$)² $\cdot$ [p1(1 – p1) + p2(1 – p2)] / (p1 – p2)²

The sample size for each group is approximately 132 cases.

This study is a retrospective study, involving the collection of data from 160 pregnant women with normal blood glucose levels and 107 pregnant women with gestational hyperglycemia. Statistical analysis revealed a significant difference in the delivery methods between the two groups. Therefore, the enrolled cases in this study consist of 107 pregnant women with gestational hyperglycemia and 160 pregnant women with normal blood glucose levels.

Statistical analyses were conducted using SPSS 26.0 software (IBM, Armonk, NY, USA). The assumptions for each test were verified prior to analysis:

Continuous Data:

(1) Normality: Assessed using the Shapiro-Wilk test (for n 50) or Kolmogorov-Smirnov test (for n $\ge$50).

(2) Homogeneity of Variances: Evaluated with Levene’s test for equality of variances.

① Parametric tests (Independent Samples t-test):

(a) Assumed normal distribution and equal variances (if Levene’s test p $>$ 0.05).

(b) Reported as mean $\pm$ standard deviation (SD).

② Non-parametric tests (Mann-Whitney U test):

(a) Applied when data were skewed (p 0.05 for normality) or variances were unequal (p 0.05 for Levene’s test).

(b) Reported as median (interquartile range, IQR).

Categorical Data:

Chi-square test:

① Assumed independence of observations and expected cell counts $\ge$5 in $\ge$80% of cells.

② Used to compare proportions between groups (e.g., delivery modes, complications).

③ Fisher’s exact test:

(a) Applied when $\ge$20% of cells had expected counts 5 to avoid Type I error inflation.

(b) Reported p-values derived from exact probabilities.

Significance Level:

All tests were two-tailed, with p 0.05 considered statistically significant.

There were no significant differences in maternal age, gestational weight gain, or the proportion of male infants between the 2 groups (p $>$ 0.05). However, the hyperglycemic group exhibited a higher pre-pregnancy BMI compared to the normoglycemic group. Furthermore, the gestational age at delivery was lower in the hyperglycemic group, while the birth weight was higher than that of the normoglycemic group. These differences were statistically significant (p 0.05), as indicated in Table 1.

The trial of labor rates did not show a significant difference between the hyperglycemic group and the normoglycemic group (p $>$ 0.05) as illustrated in Table 2. However, the cesarean section rate was notably higher in the hyperglycemic group, whereas the rate of vaginal deliveries, including those assisted by forceps, was lower compared to the normoglycemic group. These differences were statistically significant (p 0.05), as shown in Table 3. There was no statistical difference in the indications for cesarean section between the 2 groups, as shown in Table 4.

Based on the indications for cesarean section, cases involving fetal macrosomia, cephalopelvic disproportion, and arrested labor were selected from both the hyperglycemic and normoglycemic groups. These cases were then compared to vaginal delivery cases in terms of neonatal weight. In the hyperglycemic group, neonates delivered via cesarean section due to these indications had a significantly higher birth weight compared to those delivered vaginally, with a statistical result of Z = 3.500 and p 0.05. Similarly, in the normoglycemic group, neonates delivered by cesarean section for the same reasons also exhibited significantly higher birth weights than their vaginally delivered counterparts, indicated by Z = 3.750 and p 0.05. However, when examining cesarean section cases specifically for these indications, there was no significant difference in birth weight between the hyperglycemic and normoglycemic groups, as shown by Z = 1.541 and p = 0.123. Likewise, among the vaginal delivery cases, birth weight did not differ significantly between the 2 groups, with results of Z = 1.209 and p = 0.227 (Table 5).

The rates of postpartum hemorrhage, fetal distress, and neonatal asphyxia did not show significant differences between the 2 groups (p $>$ 0.05). However, the hyperglycemic group experienced higher incidences of shoulder dystocia (the denominator used in the statistics for shoulder dystocia is the number of vaginal deliveries plus the number of forceps deliveries) and intrapartum fever compared to the normal blood glucose group, with both differences being statistically significant (p 0.05), as illustrated in Table 6.

GDM has seen a significant rise in incidence globally, attributed to various risk factors including obesity, advanced maternal age, and lifestyle changes. The prevalence of GDM among women with pre-existing obesity is notably higher, with studies indicating that pre-pregnancy BMI is a critical determinant of GDM risk [12, 13]. Additionally, advanced maternal age, especially women over 35 years, has been associated with increased risk of GDM [14]. Macrosomia, defined as a birth weight greater than 4000 grams, presents a significant public health challenge. Studies indicate that the prevalence of macrosomia can range from 0.5% in India to as high as 13.9% in China [15, 16]. This variation is influenced by factors such as maternal obesity, gestational diabetes, and regional dietary habits. Notably, infants born to mothers with gestational hyperglycemia have a 2–3 times higher risk of macrosomia compared to those born to normoglycemic mothers [12]. The increasing trend of macrosomia is concerning, as it is associated with adverse perinatal outcomes, including higher rates of cesarean delivery and neonatal complications [17].

In accord with these observations, our study found that the hyperglycemic group exhibited a higher pre-pregnancy BMI and cesarean delivery rate, alongside increased neonatal birth weight and a greater incidence of intrapartum fever compared to the normal blood glucose group.

This study found no significant differences in the rates of fetal stress, neonatal asphyxia, or postpartum hemorrhage between the 2 groups. Both groups included infants classified as macrosomic, and GDM did not appear to elevate the risk of these delivery complications. Previous research indicates that the incidence of shoulder dystocia in infants weighing between 4000 g and 4500 g is approximately 5% [18]. In our study, the incidence of shoulder dystocia in the normal blood glucose group was recorded at 3.2%, aligning with existing literature. However, the hyperglycemic group exhibited a shoulder dystocia incidence of 16%, which is significantly higher than that of the normal blood glucose group. Pregnancies complicated by diabetes result in a macrosomic fetus developing a unique pattern of excessive growth characterized by central fat deposition [19]. Research by Duewel et al. [20] identified several independent risk factors for shoulder dystocia, including an estimated fetal weight of 4250 g or more, a difference of 2.5 cm or greater between abdominal circumference and head circumference, and the presence of diabetes. Additionally, another study highlighted that large-for-gestational-age infants, diabetes (both pregestational and GDM), and vacuum-assisted delivery are significant risk factors for shoulder dystocia [21]. Furthermore, a retrospective cohort study involving Chinese women found that a pre-pregnancy BMI of 24 kg/m² or higher was linked to an increased risk of GDM, shoulder dystocia, and cesarean delivery [22]. In our study, the average pre-pregnancy BMI in the hyperglycemic group was 24.6 kg/m², surpassing the 24 kg/m² threshold. Consequently, the elevated incidence of shoulder dystocia in the hyperglycemic group may be attributed to hyperglycemia, a high pre-pregnancy BMI, or a combination of both factors; further investigation is necessary to delineate their individual impacts.

This study found that while the incidence of shoulder dystocia was higher in the hyperglycemic group, there was no significant difference in the rates of neonatal asphyxia between the hyperglycemic and control groups, and notably, no cases of neonatal brachial plexus injury were reported in either group. This aligns with existing literature indicating that, over the past two decades in Sweden, the incidence of shoulder dystocia has risen, whereas the occurrence of neonatal brachial plexus palsy has actually declined [23]. Therefore, the findings from this study imply that even though shoulder dystocia is more common among those with hyperglycemia, it does not appear to elevate the risk of neonatal injuries, including asphyxia or brachial plexus injury.

In clinical practice, accurately predicting macrosomia is crucial for both maternal and fetal health. Ultrasound measurement plays a significant role in predicting macrosomia prenatally. When the abdominal circumference is $\ge$36 cm, the sensitivity of predicting macrosomia can reach 80–90%, while the specificity is about 70% [24]. Head circumference and femur length are also important ultrasound parameters. However, when using these parameters to estimate fetal weight, the error rate can be as high as 10–15% [25]. Ultrasound technology needs to be continuously updated and improved to enhance the accuracy of prediction. For instance, combining multiple ultrasound parameters with maternal clinical characteristics may help improve the accuracy and reliability of EFW (estimated fetal weight) estimation [26].

In the study of prenatal prediction of macrosomia (fetal weight $\ge$4000 grams), biomarkers play an important role, especially in identifying high-risk populations. Research shows that GDM is one of the main risk factors for macrosomia, especially when fasting blood glucose is $\ge$5.1 mmol/L, with the risk of macrosomia increasing by 2.5 to 3 times [27]. The latest research shows that, A higher TyG index (the triglyceride-glucose index) and a higher TG/HDL-C (triglyceride to high-density lipoprotein cholesterol) ratio in maternal blood in late gestation indicate a metabolic process that causes fetal macrosomia. Physicians should be more cautious about the risk of macrosomic fetuses when the cut-off value for the TyG index ($>$4.88) and the cut-off value for the TG/HDL-C ratio ($>$3.63) are available [28].

The indication for cesarean section in pregnant patients, whether they have gestational hyperglycemia or normal glucose levels, continues to raise concerns. Research indicates that fetal weight plays a significant role in the likelihood of cesarean delivery, especially in women diagnosed with GDM. Current findings suggest that a fetal weight of 4150 to 4190 grams may be a reasonable threshold for considering a cesarean section in both groups of pregnant patients. However, it is essential to take into account individual pelvic findings and the mother’s preferences regarding the mode of delivery when making specific clinical decisions. Additionally, research indicates that the risk of cesarean delivery increases notably in cases of fetal macrosomia, defined as a fetal weight of 4500 grams or more, particularly in pregnancies complicated by diabetes [29]. To mitigate cesarean section rates and minimize the risk of neonatal birth trauma, current evidence underscores the importance of maintaining strict glycemic control and creating individualized delivery plans [30].

In conclusion, while we suggest a weight range of 4150–4190 g as a guideline for considering a cesarean section, it is essential that the decision-making process incorporates a thorough clinical assessment and takes into account the unique circumstances of each pregnant woman. Additionally, future studies should further investigate how varying weight thresholds correlate with cesarean section rates, which will enhance our ability to effectively inform clinical practice.

Our retrospective design may introduce selection bias. Additionally, the generalizability of our findings to multiparous women or non-Asian populations is uncertain. Strengths include the large sample size and detailed ascertainment of perinatal complications using standardized definitions.