Association between oral glucose tolerance test (OGTT) glucose levels and fetal macrosomia in non-diabetic women

Introduction

Foetal macrosomia is defined by the American College of Obstetricians and Gynaecologists (ACOG) as an estimated foetal weight exceeding 4,000 or 4,500 g, regardless of gestational age [1]. Another definition of excessive foetal growth is large for gestational age (LGA), which refers to an estimated foetal weight above the 90th percentile for a given gestational week [1].

Foetal macrosomia accounts for approximately 7 % of all pregnancies and is associated with both maternal and foetal/neonatal complications [2]. Maternal complications include labour arrest, increased rates of emergency caesarean delivery, uterine atony, postpartum haemorrhage, deep perineal lacerations, and anal sphincter injury [3], 4]. Foetal and neonatal complications may include shoulder dystocia and its sequelae, such as brachial plexus injury, clavicle and humerus fractures, and foetal asphyxia. While the incidence of shoulder dystocia is approximately 1 % in deliveries with birth weights less than 4,000 g, this risk increases to 5–10 % when the birth weight exceeds 4,000 g [5]. In addition to these immediate perinatal complications, foetal macrosomia is associated with long-term metabolic consequences. Newborns with macrosomia are at increased risk for childhood obesity, insulin resistance, and type 2 diabetes later in life, highlighting the importance of early identification and management of maternal hyperglycaemia [1], 6].

Foetal gluconeogenesis is limited in healthy pregnancies, and the primary source of glucose in the foetal circulation is maternal glucose, which is transferred via the placenta through the GLUT-1 transporter on the basis of the maternal–foetal glucose concentration gradient. An increase in the amount of glucose transferred to the foetus stimulates foetal insulin production, which enhances glucose uptake by tissues and leads to fat accumulation in adipose tissue, ultimately contributing to foetal macrosomia. Additionally, increased levels of human placental lactogen (HPL), cortisol, insulin-like growth factor, growth hormone, and other growth factors support the development of macrosomia [7]. In addition to maternal diabetes, other contributing factors, such as maternal obesity, grand multiparity, excessive gestational weight gain, and advanced maternal age, have been associated with foetal macrosomia. Moreover, placental size, GLUT-1 expression, and the amount of glucose utilized in placental metabolism influence the quantity of glucose reaching the foetus, thereby impacting the development of macrosomia [8].

Oral glucose tolerance tests (OGTTs) are typically performed between the 24th and 28th weeks of pregnancy to screen for gestational diabetes mellitus (GDM). The OGTT can be conducted as a one-step procedure with 75 g of glucose solution or as a two-step approach involving an initial 50 g screen followed by a 100 g diagnostic test. While the American Diabetes Association (ADA) recommends the one-step method [9], ACOG advocates the two-step approach [10]. A 75 g glucose solution is administered after an overnight fast in the one-step method, and glucose levels are measured at fasting, the first hour, and the second hour. GDM is diagnosed if any one of these values exceeds the defined thresholds. Owing to its simplicity, the 75 g OGTT is more widely used and has been shown in several studies to yield results comparable to those of the two-step approach, particularly in predicting foetal macrosomia [11].

Many studies have focused on the relationship between OGTT glucose levels and foetal macrosomia in women diagnosed with GDM [12]. However, numerous macrosomic infants are born to mothers without GDM or impaired glucose tolerance. Identifying this subgroup without apparent risk factors is crucial for preventing foetal complications.

The aim of this study was to evaluate the relationships between fasting, first-hour, and second-hour glucose values from the 75 g OGTT and the risk of foetal macrosomia in non-diabetic (non-GDM) pregnancies to determine which parameter most strongly predicts macrosomia.

Materials and methods

This cross-sectional study was conducted at the Basaksehir Cam and Sakura City Hospital Obstetrics and Gynaecology Department between September 1, 2023, and September 1, 2024. This research was granted ethical approval by the Ethics Committee of Basaksehir Cam and Sakura City Hospital on November 21, 2023, with protocol number KAEK/11.10.2023.480 and was conducted in compliance with the Declaration of Helsinki.

Pregnant women who attended routine antenatal visits at 24–28 weeks of gestation and underwent a standard 75 g OGTT were enrolled in the study. In our setting, the 75 g OGTT is routinely recommended for all pregnant women between 24 and 28 weeks of gestation as part of national antenatal care guidelines, irrespective of individual risk factors. The demographic data and OGTT results of the participants were recorded at this stage. After delivery, additional data were retrospectively collected from the birth reports, including GDM diagnosis status, type of treatment in diagnosed cases, gestational age at delivery, mode of delivery, occurrence of delivery-related complications, and neonatal birth weight.

The inclusion criteria were pregnant women aged 16–49 years who presented to the antenatal clinic during 24–28 weeks of gestation and consented to undergo an OGTT as recommended by their physicians. The exclusion criteria were a prior diagnosis of diabetes mellitus (DM), multiple pregnancies, or placental insufficiency disorders such as preeclampsia and foetal growth restriction (FGR). Patients who delivered preterm, who delivered at another institution, or who developed preeclampsia or FGR later in pregnancy (despite no diagnosis at the time of initial data collection) were also excluded. Women diagnosed with GDM following the OGTT were recorded as a separate subgroup for descriptive analysis. However, they were not included in the primary comparison between macrosomic and normal-weight foetuses, as the aim of the study was to assess the relationship between OGTT parameters and foetal macrosomia in nondiabetic women, thereby avoiding potential confounding related to abnormal glucose metabolism.

Patients diagnosed with GDM on the basis of their OGTT results were evaluated as a separate subgroup. All patients included in the final analysis were routinely followed at our tertiary centre with monthly antenatal visits. During these visits, fasting and postprandial blood glucose levels were checked, and women with normal OGTT results maintained normal glucose levels throughout follow-up until delivery. All 54 patients diagnosed with GDM received their diagnosis on abnormal OGTT results. On the basis of the results of the Hyperglycaemia and Adverse Pregnancy Outcome (HAPO) Study published in 2011, diagnostic criteria were established by the International Association of Diabetes and Pregnancy Study Groups (IADPSG) [13]. These criteria are currently recommended by the ADA. Accordingly, the diagnostic thresholds are defined as a fasting plasma glucose level of ≥92 mg/dL, a first-hour value of ≥180 mg/dL, and a second-hour value of ≥153 mg/dL. The diagnosis of GDM can be made if any one of these values is abnormal. Foetal macrosomia was defined as a birth weight ≥4,000 g.

The required sample size was calculated using G*Power for MacOS version 3.1. Assuming an estimated incidence of macrosomia of 10 %, with a Type I error rate (α) of 0.05, effect size of 0.5, and 95 % confidence level, the minimum sample size required to achieve 80 % power was determined to be 246 participants.

Statistical analysis of the data was performed using IBM SPSS Statistics for MacOS, version 20.0. The normality of distribution for continuous variables was assessed using the Shapiro–Wilk test. Continuous variables are expressed as the mean ± standard deviation for normally distributed data and as the median (Q1–Q3) for nonnormally distributed data. Categorical (nominal) variables are presented as frequencies (percentages). Differences in means between groups were analysed using Student’s t-test for normally distributed data and the Mann–Whitney U test for nonnormally distributed data. Associations between categorical variables were assessed using the chi-square test. Receiver operating characteristic (ROC) curve analysis was performed to evaluate the predictive value of the second-hour OGTT glucose levels. A p value of <0.05 was considered statistically significant.

Results

The study flowchart is presented in Figure 1. A total of 1,047 patients for whom OGTTs were recommended were initially screened. A total of 267 patients who declined to participate, 239 patients who did not complete the OGTT (due to intolerance to the glucose solution or failure to provide blood samples during the first and second hours), and 24 patients with multiple pregnancies were excluded from the study. Among the remaining 517 patients, 4 patients who were later diagnosed with FGR and 3 patients who were diagnosed with preeclampsia after the test request week, as well as 66 patients who delivered preterm and 205 patients who were lost to follow-up at our hospital, were also excluded. Of the remaining 239 patients, 54 were diagnosed with GDM and were included in the analysis of demographic characteristics and subsequently subjected to subgroup analysis. After excluding patients diagnosed with GDM, the remaining 185 patients were categorized as follows: those who delivered macrosomic foetuses constituted Group 1 (n=16), and those who delivered normal birth weight foetuses comprised Group 2 (n=169).

Figure 1:

Flowchart of the study.

Patients diagnosed with GDM – defined as having at least one abnormal value among fasting, first-hour, or second-hour OGTT glucose levels – were excluded from the main analysis for further subgroup analysis. The remaining patients with normal OGTT values were divided into two groups: those with macrosomic newborns (Group 1, n=16) and those with newborns of normal birth weight (Group 2, n=169). The groups were compared in terms of maternal age, gravidity, parity, gestational week (weeks + days), OGTT fasting, first-hour, and second-hour glucose levels (mg/dL), body mass index (BMI), gestational week at delivery (weeks + days), neonatal birth weight, and first- and fifth-minute APGAR scores using Student’s t-test (Table 1). The presence of comorbidities, foetal anomalies, and mode of delivery (vaginal vs. caesarean section) were compared between the groups using the chi-square test (Table 1). There were no statistically significant differences between Groups 1 and 2 in terms of maternal age, gravidity, parity, gestational age, BMI, delivery week, APGAR score at the first or fifth minute, presence of comorbidities, foetal anomaly diagnosis, or mode of delivery (vaginal and caesarean section) (p=0.397, p=0.349, p=0.156, p=0.154, p=0.061, p=0.074, p=0.247, p=0.664, p=0.586, p=0.775, p=0.477, p=0.477, respectively). Similarly, fasting and first-hour OGTT glucose values did not significantly differ between the two groups (p=0.425 and p=0.429, respectively).

The mean birth weight of the newborns in Group 1 was significantly higher than that in Group 2 (p=0.000). Additionally, the mean second-hour OGTT glucose concentration was significantly higher in Group 1 (104.38 ± 15.84 mg/dL) than in Group 2 (95.19 ± 17.40 mg/dL) (p=0.041).

A ROC analysis was performed to evaluate the predictive value of the second-hour OGTT glucose level in distinguishing between the macrosomic and normal birth weight groups (Figure 2). The area under the curve (AUC) was calculated as 0.675, with a p value of 0.021, indicating a statistically significant discriminatory ability (Table 2). The optimal cut-off value was determined to be 99.5 mg/dL based on the point at which the sensitivity and specificity were closest. Accordingly, a second-hour OGTT glucose concentration above 99.5 mg/dL predicted foetal macrosomia, with a sensitivity of 62.5 % and a specificity of 63.3 %.

Figure 2:

Receiver operating characteristic (ROC) curve for the oral glucose tolerance test (OGTT) second-hour value.

A total of 54 patients diagnosed with GDM were excluded from the main analysis and are presented for descriptive purposes only. The mean maternal BMI in this group was 30.48 ± 4.87 kg/m². Among the GDM patients, 15 (27.8 %) were managed with dietary modification alone, while 39 (72.2 %) required insulin therapy. The mean fasting, first-hour, and second-hour OGTT glucose levels were 90.1, 185.4, and 145.6 mg/dL, respectively. Despite the high-risk metabolic profile, the mean neonatal birth weight was 3,212 ± 514.7 g. The caesarean delivery rate was 75.9 % (n=41). Owing to the small number of macrosomic cases in this subgroup (n=3), no further subgroup analysis was performed.

Discussion

Early diagnosis and prevention of foetal macrosomia are important because of serious maternal/neonatal complications. Modifiable risk factors such as excessive gestational weight gain, maternal obesity, and diabetes are important for preventing macrosomia early. Recent large-scale studies have also proposed multivariable prediction models to better identify pregnancies at risk of macrosomia [14]. Luo et al. reported that a maternal height ≥165 cm, prepregnancy overweight, and at least two abnormal glucose values on the 75 g OGTT were significant predictors of macrosomia [14]. In our study, we aimed to investigate the relationships between foetal macrosomia and fasting, first-hour, and second-hour glucose values obtained from the 75 g OGTT in non-GDM patients.

Khan et al. examined the association between foetal macrosomia and second-hour glucose levels after a 75 g OGTT was performed at 16–20 weeks of gestation in 1,331 pregnant women without GDM [15]. The results were stratified into five groups based on second-hour glucose levels, and a linear increase in macrosomia incidence was observed with increasing glucose levels. While the incidence was 1.2 % in the lowest glucose group, it reached 9.5 % in the highest group. A significant linear correlation was found between glucose levels and birth weight. The study concluded that even mild hyperglycaemia in women without GDM increases the risk of macrosomia. Notably, the OGTT in that study was performed at 16–20 weeks of gestation. Abnormal OGTT results at this gestational age may indicate pregestational diabetes according to current diagnostic criteria. Additionally, the lack of follow-up for glucose metabolism in later pregnancy weeks should be considered, as undiagnosed GDM developing after the OGTT may have contributed to the observed macrosomia. In contrast, our study evaluated 75 g OGTT results in patients without a diagnosis of GDM or pregestational diabetes and differs from the aforementioned study in terms of the gestational timing of the OGTT as well as the exclusion of patients who were later diagnosed with GDM.

Geifman-Holtzman et al. investigated the relationship between foetal macrosomia and the results of a 100 g, 3 h OGTT performed at term in women who had previously tested negative on the 50 g OGTT during 24–28 weeks of gestation but were suspected of having macrosomia at term (37–40 weeks) [16]. The study included 170 term pregnancies with an estimated foetal weight above the 90th percentile (or >4,000 g). Impaired glucose metabolism was detected in 10 participants (5.9 %) at term. There was no statistically significant difference in mean foetal weight between patients with normal and abnormal 100 g OGTT results, and no association was found between macrosomia and abnormal OGTT values at term. Although OGTTs at term did not increase the detection rate of macrosomia, the authors suggest that they may still influence patient management and improve obstetric outcomes. Unlike our study, this investigation did not evaluate the numerical association between specific OGTT values and foetal weight.

The study most comparable to our study in the literature was conducted by Brankica et al. in which the relationships between fasting, first-hour, and second-hour glucose values from the 75 g OGTT and foetal macrosomia were investigated [12]. The study included 118 patients, 78 of whom were diagnosed with GDM. The incidence of macrosomia was significantly higher in patients with GDM than in those without GDM (30.7 vs. 5.0 %). Gestational age at delivery and fasting glucose levels were identified as independent predictors of macrosomia. Among the OGTT values, fasting glucose had the highest predictive power for macrosomia (AUC: fasting 0.782; first-hour 0.719; second-hour 0.51). The notably high proportion of GDM cases in that study (66.1 %) suggests a greater likelihood of extreme OGTT values. In contrast, our study excluded patients with GDM to achieve a more homogeneous distribution of fasting, first-hour, and second-hour glucose values. It is more likely that macrosomia can be diagnosed and managed earlier in GDM patients since they are often monitored closely for macrosomia and managed through dietary or medical interventions. However, in patients without a diagnosis of GDM and without evident risk factors, macrosomia may go unnoticed. Therefore, early identification of macrosomia in this population is particularly important.

The relationship between 50 g of OGTT glucose and foetal macrosomia was evaluated in 980 pregnant women at 24–28 weeks of gestation, excluding those diagnosed with GDM [17]. Participants were categorized into three groups based on their 50 g OGTT results: those with first-hour plasma glucose levels ≥140 mg/dL but normal 100 g OGTT results (false-positive group), those with levels between 130 and 139 mg/dL, and those with levels <130 mg/dL (control group). No statistically significant differences were found between the groups in terms of foetal weight, macrosomia, or LGA rates. The incidence of macrosomia was 1.25 % in the false-positive group, 1.07 % in the 130–139 mg/dL group, and 1.1 % in the <130 mg/dL group; the LGA rates were 13.1 %, 13.5%, and 12.9 %, respectively. The results of this study revealed that the 50 g OGTT has limited predictive value for macrosomia and LGA among pregnant patients without a GDM diagnosis.

In a retrospective study by Bai et al. the records of 998 patients diagnosed with GDM were analysed to evaluate the impact of different GDM subgroups on the incidence of macrosomia (classified based on 75 g OGTT results) [18]. Seven subgroups were defined according to combinations of elevated fasting, first-hour, and second-hour glucose values. The highest incidence of macrosomia (21.92 %) was observed in Group C, where all three glucose values were elevated. This study provided a more detailed analysis of the known association between GDM and macrosomia, highlighting that patients with elevations in all three glucose measurements are at greater risk. However, a notable limitation of the study is the lack of information regarding treatment or glucose monitoring during the follow-up period.

A review of the literature reveals that our study differs from others by including patients with normal glucose values on the 75 g OGTT (excluding those with GDM) and aiming to identify a predictive value for macrosomia through comparison of OGTT results between macrosomic and normal foetal weight groups. In our study, patients with GDM were excluded to minimize the confounding effect of abnormal glucose metabolism on foetal macrosomia.

One of the strengths of our study is the homogeneity between Groups 1 and 2 in terms of parameters that may contribute to foetal macrosomia, such as BMI, gravidity, parity, maternal age, and gestational age at delivery, with no statistically significant differences observed (p>0.05). Another advantage is the separate analysis of patients diagnosed with GDM after applying exclusion criteria, thereby eliminating the confounding effect of abnormal glucose metabolism on macrosomia. Owing to the presence of only three macrosomic foetuses within the GDM subgroup, a comparative analysis between macrosomic and normal-weight foetuses was not performed. The low incidence of macrosomia in the GDM group can be attributed to the effective glycaemic control achieved through dietary or medical treatment following the GDM diagnosis via OGTT.

The analysed sample (n=239) was slightly smaller than the planned 246 participants because of loss to follow-up during the data collection period and later exclusion of more than expected patients; however, the findings remain robust. Although no statistically significant differences were observed between Groups 1 and 2 for maternal BMI, age, or parity, it should be noted that the relatively small number of macrosomic cases (n=16) may have limited the statistical power to detect such differences. These factors could still be clinically relevant in a larger cohort. An important limitation of this study is the imbalance between groups, with a relatively small number of macrosomic cases. While this may limit statistical power and multivariable adjustment, the observed associations suggest a potential relationship between OGTT second-hour glucose levels and fetal macrosomia, which may be confirmed in larger cohorts. The area under the curve (AUC) from the ROC analysis in our study was 0.675; expanding the sample size may increase this value closer to 1, thus enhancing the study’s power. Additionally, the ROC analysis showed fair discrimination (AUC 0.675). Due to the limited number of cases, the identified cut-off of 99.5 mg/dL should be considered exploratory rather than definitive. Finally, the baseline BMI is presented in the study; however, it is known that maternal weight gain during pregnancy is an important risk factor for macrosomia. Since weight gain data were missing for some of the patients, a robust analysis was not possible.

Macrosomia may remain unpredictable in the absence of obvious risk factors such as GDM, maternal obesity, excessive gestational weight gain, a history of macrosomic birth, or a history of maternal macrosomia. Consequently, obstetric follow-up and delivery planning might not anticipate macrosomia, potentially resulting in more catastrophic outcomes. In our study, second-hour glucose values from the 75 g OGTT exceeding 99.5 mg/dL were found to be associated with macrosomia and may serve as valuable predictors. These findings should be supported by multicentre studies with larger sample sizes.