Preterm birth is defined as birth that occurs before the 37th week of pregnancy, and its prevalence varies worldwide, ranging from 4% to 16% [1, 2]. More than 13 million babies are born prematurely each year, before the 37th week of pregnancy, which equates to almost 1 in 10 babies being born preterm [3]. Preterm birth and its associated complications are leading causes of death in children under five years old, with nearly 1 million children dying annually as a result of these complications [4]. Many major health issues affect survivors, such as diabetes, hypertension, ischemic heart disease, and chronic renal disease [5, 6, 7]. Therefore, the prediction, prompt intervention, and prevention of preterm birth are crucial. Preterm birth is characterized as a syndrome with multiple contributing variables. As a result, there is limited effectiveness in the techniques employed to predict this complex condition. Early hypothalamic-pituitary-adrenal (HPA) axis activation has received more attention among the variables contributing to preterm birth [8]. Lower progesterone activity and higher concentrations of prostaglandin, protease, and proinflammatory cytokine are known to be higher when the HPA axis is activated early [9]. The development of the adrenal gland, a key element of the HPA axis, occurs in the early stages of embryogenesis. The cortex of the adrenal gland differentiates into the fetal zone (FZ), transitional zone (TZ), and definitive zone (DZ) starting from the 9th week of pregnancy. Of these, the FZ is primarily composed of large cells (20–50 mm) with steroidogenic activity. The size of the entire adrenal gland and the internal FZ in fetuses can significantly increase due to the stressful environment caused by inflammation [10, 11]. Sonographic evaluation of the fetal adrenal gland has revealed that it serves as a marker of placental insufficiency and preterm birth, yet these findings have not been integrated into clinical practice [12, 13]. Another technique for predicting preterm birth is cervical length (CL) measurement, which is now included in all screening programs [14, 15]. CL is measured according to stringent guidelines, with measurements influenced by maternal characteristics, image quality, and the operators’ experience. According to Iams et al. [16], 20% of sonographers who applied for certification were denied on their first attempt. However, preterm birth can be predicted by measuring the fetal adrenal gland, which has a size similar to that of the fetal kidneys from the second trimester onward, and the form and borders of the adrenal can be easily visualized ultrasonographically. The aim of this study was to determine whether the FZ and volumetric measures of the fetal adrenal glands in term and preterm fetuses are suitable for use in clinical settings as indicators of preterm birth.

A total of 238 women were included in the study, with 220 (92.4%) in the term group and 18 in the preterm group. The groups were comparable in terms of age, body mass index (BMI), gestational week at the time of fetal adrenal volume measurement, fetal weight, gravidity, parity, and abortion (Table 1).

Although the AGV (p = 0.039), FZV (p = 0.001), cAGV (p 0.001), and cFZV (p 0.001) were greater and statistically significant in the preterm group than in the term group, no significant differences were found between the groups in terms of the FZV/AGV ratio (p = 0.224) or CL (p = 0.540) (Table 2).

ROC curve analysis revealed that patients with an AGV of $\ge$726.56 mm${}^3$ and an FZV of $\ge$538.90 mm${}^3$ were at an increased risk for preterm birth. However, these parameters demonstrated weak discriminatory performance, with area under the curve (AUC) values for identifying preterm birth of 0.693 and 0.647, respectively (Table 3, Fig. 2).

When considering the corrected values, the AUC values for cAGV, and cFZV in predicting preterm birth were 0.899 and 0.868, respectively. The optimal cutoff values of cAGV and cFZV for the differentiating preterm birth were $\ge$1.012 mm${}^3$/kg, and $\ge$0.493 mm${}^3$/kg, respectively (Table 3, Fig. 2).

Although not statistically significantly, when a CL of $\le$35 mm [19] was considered a risk factor for preterm birth, a total of 57 (23.9%) patients had a CL $\le$35 mm, with 49 (22.7%) in the term group and 8 (44.4%) in the preterm group. The percentage of patients with a CL $\le$35 mm was higher in the preterm group compared to the term group (44.4% vs 22.7%, p = 0.067).

When the CL was $\le$35 mm and the cutoff values of AGV, FZV, cAGV, and cFZV determined in the present study were combined, the percentages of patients in the term and preterm groups, as well as the sensitivity and specificity values, were calculated, as shown in Table 4.

Considering both length and volume together predicted fewer premature births in all patients. Therefore, the sensitivity values for all parameters decreased remarkably, while the specificity values increased. The most significant increase in specificity was observed for AGV ($\ge$726.56 mm${}^3$). In summary, cAGV and cFZV appeared to be good predictors of preterm birth, both when considered individually in combination with CL (Table 4).

This case-control study aimed to evaluate the effectiveness of a simple approach that obstetricians may include into routine practice, either as a supplement to or in combination with existing techniques for predicting preterm birth. The findings demonstrated that two useful markers for predicting preterm birth are elevated AGV and increased volume of the FZ, a significant component of the fetal adrenal gland. Our proposed strategy for predicting preterm birth may not be widely adopted, as most obstetricians in many underdeveloped countries lack the capability to perform 3D sonographic scans, despite the increasing availability of advanced ultrasound technology.

During fetal development, the adrenal gland consists of three regions: the DZ, TZ, and FZ. The FZ exhibits the highest steroidogenic activity among the three adrenal gland zones and secretes significant amounts of dehydroepiandrosterone (DHEA) and DHEA-sulphate. These hormones contribute to the production of placental estrogen, which plays a crucial role in initiating labor.

Turan et al. [20] evaluated AGV in the prediction of preterm birth and reported that increased AGV was associated with preterm birth. In the present study, results showed that the AGV was greater in the preterm birth group. However, when comparing the FZV, where steroidogenic activity was higher, the difference between the term and preterm groups was even more pronounced.

Hoffman et al. [18] found that AGV was lower in the preterm birth group compared to the term birth group when investigating AGV in asymptomatic pregnant women in the third-trimester of pregnancy. Our findings, however, revealed the opposite outcomes. The group with preterm births presented higher FZV and AGV.

Hoffman et al. [12] found that the FZV/AGV ratio did no predict preterm birth. Consistent with this finding, our study also showed that the FZV/AGV ratio was similar in women with term and preterm birth.

Currently, World Association Perinatal Medicine recommends a stepwise approach for predicting preterm birth. This method can significantly reduce unnecessary treatments. According to this approach, if the CL 15 mm, admission and treatment are advised. If the CL $>$30 mm, reassurance is offered. For a CL between 15 and 30 mm, biomarker testing is recommended [21]. Hence, in our study, patients with CL measurements $\le$30 mm were informed about the risk of preterm birth, and their follow-up and treatment were planned accordingly. Akar et al. [17] in a study conducted in Turkiye with a population similar to ours, reported that the 10th percentile of CL at 24 weeks in singleton pregnancies was equivalent to 30 mm, and the 50th percentile was equivalent to 37 mm. Close follow-up and treatment were recommended, if necessary, to prevent preterm birth in patients with CL below the 10th percentile [17]. However, there is a significant uncertainty, as many patients fall between the CL measurement of the 10th and 50th percentiles. In our study, 23% of the individuals exhibited CLs between 30 and 35 mm. Additionally, the CLs of the preterm and term groups at 24–28 weeks were similar.

There are many factors that can cause preterm birth. It is still unclear whether alterations in fetal AGV and FZ volume are caused by external variables or if they could initiate a chain of events leading to premature birth. However, Agarwal et al. [22] evaluated women at 28–37 weeks of gestation and reported that both CL measurement and increased FZV are valuable methods that can be used for predicting preterm birth in women at risk.

In the current study, we assessed pregnancies between 24 and 28 weeks of gestation and waited to take another measurement until after the baby was born. As a result, one of the shortcomings of our study was that we did not assess changes in the AGV or FZ volume in pregnant patients after 28 weeks. Other limitations of our study include the possibility for unidentified confounding variables that may have affected the outcome. A further limitation was the small number of patients in the preterm birth group, accounting for only 7.6% of the whole cohort.

Our study only investigates the relationship between AGV and FZV and preterm birth. We found that both are valuable in preterm birth prediction. In our study, we found no significant in CLs between the preterm and term groups at 24–28 weeks, suggesting that AGV and FZV may provide earlier prediction of preterm birth. However, due to the small number of preterm cases in our study, further research with larger cohorts of preterm births is needed to prove this hypothesis.

To the best of our knowledge, no published study has combined fetal AGVs and CL measurements between 24 and 28 weeks for the prediction of preterm birth. The results of our study showed that fetal cAGV and cFZV appeared to be good predictors of preterm birth, whether considered individually or in combination with CL.

The study results indicated that 3D assessment of AGVs between 24 and 28 weeks of gestation, combined with CL measurement, provides a reliable and non-invasive diagnostic tool. This method can be used by obstetricians in daily practice to further investigate cases where there is a risk of premature birth.

The authors of the study are custodians of the data in anonymous form, which can possibly be provided to anyone who makes a motivated and reasoned request.

Conceptualization, AB, BK, SH, RS and AEK; methodology, AB, BK, AEK, SH, RS, AT and MHD; formal analysis, MAS; investigation, AB, BK; data curation, AB, BK, FGG and EY; writing—original draft preparation, AB, BK, AT; writing—review and editing, AB, BK, EY, AEK, RS, AT, SH, FGG, MAS and MHD; visualization, EY, BK; supervision, AT, RS, MHD, MAS, FGG and SH; project administration, AB. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

The local Institutional Review Board Statement (Duzce University) approved the study with the following number: 20-1622. The study was conducted by the Declaration of Helsinki. Informed consent was obtained from patients.

Not applicable.

This research received no external funding.

The authors declare no conflicts of interest. Michael H. Dahan is serving as Editor-in-Chief and one of the Guest editors of this journal. Andrea Tinelli is serving as one of the Editorial Board members and Guest editors of this journal. We declare that Andrea Tinelli and Michael H. Dahan had no involvement in the peer review of this article and has no access to information regarding its peer review. Full responsibility for the editorial process for this article was delegated to Paolo Ivo Cavoretto and George Daskalakis.