Clinical characteristics of patients with gestational diabetes mellitus depending on the type of treatment (a retrospective study)

Introduction

Gestational diabetes mellitus (GDM) is one of the most common complications of pregnancy and at the same time one of the most controversial problems. Hyperglycemia, which emerges for the first time during pregnancy¹, has been known for over 50 years, but there is still controversy about diagnostic methods and glycemic thresholds for establishing a diagnosis and initiating treatment. Most societies adhere to the recommendations of the International Association of Diabetes and Pregnancy Study Groups (IADPSG) regarding the need for screening on carbohydrate metabolism disorders and the threshold values in pregnant women for diagnosing GDM using a two-hour oral glucose tolerance test (OGTT) with 75 g glucose (5.1/10.0/8.5 mmol/l). The American College of Obstetricians and Gynecologists admits the use of this method but prefers a two-stage screening, which is performed for all pregnant women in several stages and with higher threshold values within the three-hour test, namely 5.3/10.0/8.6/7.8 mmol/l. The Russian Association of Endocrinologists adopted the IADPSG criteria back in 2012, which was also reflected in their joint project with the Russian Society of Obstetricians and Gynecologists on clinical guidelines for the management of patients with GDM in 2024² [2].

The objectivity of the indicator "Prevalence of GDM" is determined primarily by the widespread diagnostics. Among the most serious complications of GDM are preeclampsia, stillbirth, macrosomia, neonatal hypoglycemia and hyperbilirubinemia, cesarean section, and a number of others³. It has been shown that achieving targeted glucose values, namely less than 5.1 mmol/l on an empty stomach and less than 7.0 mmol/l one hour after the start of food intake, and maintaining them during pregnancy can reduce the risk of complications⁴. This is the main goal of GDM treatment. Currently, more and more studies focus on the heterogeneity of GDM and distinguish its various subtypes, including GDM with predominant insulin resistance, GDM with impaired insulin secretion by B cells, and a mixed variant of this condition [3–6]. However, despite such assumptions, the therapeutic approach to patients remains the same and does not depend on the probable subtype of GDM. When hyperglycemia is detected for the first time during pregnancy, therapy consists of prescribing non-drug treatment based on diet therapy, physical activity, and self-monitoring, and if the above measures are ineffective, drug treatment (insulin therapy) is added after 1–2 weeks. However, the currently available data on the existence of various GDM subtypes suggest that this approach can be improved or, more precisely, personalized on the basis of the pathogenetic subtype of the disease in a particular patient; that will subsequently ensure a reduction in the risks of pregnancy complications. The work described below is a part of a study devoted to the investigation of GDM subtypes, namely it is a retrospective analysis of the case histories of patients treated using various modes aimed at hyperglycemia correction in GDM. In most of the previous works, the main approach to identifying GDM subtypes was associated with the assessment of the Matsuda and Stumwoll insulin resistance indices [3–6]. However, these studies do not answer the main question of whether the treatment regimens for these patients differed. With consideration of the retrospective nature of the present study and missing the chance of calculating insulin resistance indices, to identify GDM subtypes we investigate the clinical characteristics of the patients, including phenotypic and biochemical differences, as well as possible pregnancy complications depending on the treatment regimens. Identifying these differences will help to form an insight into the clinical significance of various GDM subtypes, the verification of which may be of extreme practical importance.

The aim of the study was to evaluate the clinical characteristics of patients with GDM depending on the treatment regimens to confirm the existence of different GDM subtypes.

Materials and methods

The study was conducted in the women’s consultation centers of the State Budgetary Institution of the Rostov Region "City Clinical Hospital No. 20", the State Budgetary Institution of the Rostov Region "City Hospital No. 6", and the State Budgetary Institution of the Rostov Region "City Hospital No. 7" in Rostov-on-Don.

A retrospective analysis was conducted for medical records of 474 pregnant women who successfully delivered from 2018 to 2023, aged 18 years and older, regardless of GDM risk factors; the average age was 30.2 (27–33) years. In all pregnant women included in the study, the diagnosis of GDM was confirmed on the basis of the fasting glucose level at the term on or before the 23rd week or due to the OGTT with 75 g of glucose at the term between the 24th and 32nd weeks of pregnancy.

Depending on the treatment method received by patients, the following groups were formed: Group I included 353 pregnant women with GDM who received only diet therapy, and Group II included 121 pregnant women with GDM who used both diet and insulin therapy.

During the study, objective patient data such as age, blood pressure, weight, and BMI in each trimester and before delivery were assessed. Special attention was paid to the time points of GDM detection and crossover to insulin. We also conducted an analysis of the medications used for GDM-associated conditions, including the use of aspirin, progesterone preparations, and various multivitamins.

In addition, the biochemical blood test, general blood test, and coagulogram parameters were analyzed at the 20th and 30th weeks of pregnancy. When analyzing the data of instrumental examination methods through the ultrasound, the weight and size of the fetus were estimated depending on the trimester and immediately before delivery. The detection of concomitant diseases, such as overweight and obesity, anemic syndrome, hypothyroidism syndrome, and arterial hypertension, was performed separately. The term and type of delivery, as well as the weight of the child at birth, were also analyzed. Besides, special attention was paid to emerging complications such as preeclampsia and neonatal hypoglycemia.

Statistical analysis of the study results was performed using R (version 3.2, R Foundation for Statistical Computing, Vienna, Austria). A comparison of quantitative indicators between groups was performed using the Mann-Whitney test, and frequencies were assessed using the Fisher exact test. Differences were considered statistically significant at the level of p < 0.05. Data are presented as medians and interquartile ranges of quantitative indicators in groups.

Results

The following results obtained in the course of the study are presented below. According to the fasting glucose level in patients on or before the 23rd week of pregnancy, the diagnosis of "GDM" was confirmed in 254 pregnant women (53.6%); while according to the results of OGTT with 75 g of glucose in patients at the terms from the 24th to the 32nd week of pregnancy, GDM was verified in 220 patients (46.4%). It should be noted that statistically significant differences in the time point of GDM detection among patients from different groups were not found (p = 0.17).

Analysis of objective patient data, including age, blood pressure, weight, and BMI in each trimester and before delivery, did not find statistically significant differences. Comparative characteristics of objective data of patients with GDM are presented in Table 1.

Results of the assessment of the biochemical blood test parameters at the 30th week of pregnancy revealed higher levels of urea (p = 0.009), AST (p = 0.03), and ALT (p = 0.02) in Group II. The coagulogram results showed higher levels of prothrombin (p = 0.02) and D-dimer (p = 0.047) also in Group II. It should be noted that, despite the statistically significant increase in the examined parameters, their values did not go beyond the reference physiology values. At the same time, at the 20th week of pregnancy, no statistically significant differences were found between the groups according to the data of biochemical blood test, clinical blood test, and coagulogram. Comparative characteristics of laboratory parameters of patients with GDM are presented in Table 2.

When analyzing the size and weight of the fetus according to ultrasound examination data, it turned out that in the 1st trimester, a higher fetal weight indicator was revealed in group I (p = 0.03), while in the 2nd and 3rd trimesters, as well as immediately before delivery, the fetal weight indicators did not differ statistically significantly between groups. Comparative characteristics of fetal weight according to ultrasound data by trimesters are presented in Table 3.

Assessment of the pharmacological anamnesis revealed that patients from Group II took progesterone preparations significantly more often [ 61 (17%) patients versus 32 (26%), p = 0.02]. It should be noted that pregnant women administered Duphaston [ 28 (8%) patients in Group I and 14 (12%) women in Group II, p = 0.66] and Utrozhestan [ 29 (8%) patients in Group I and 18 (15%) women in Group II, p = 0.66]. No statistically significant differences were also found between the groups on other medications taken by patients. Comparative characteristics of medication taken by patients with GDM are presented in Table 4.

Investigation of the co-morbidities, such as overweight and obesity, anemic syndrome, hypothyroidism syndrome, arterial hypertension, as well as complications, including preeclampsia and neonatal hypoglycemia, did not reveal statistically significant differences between the groups in these features. Comparative characteristics of the concomitant pathologies and complications of GDM in patients are presented in Table 5.

It should be especially noted that in Group II, delivery occurred earlier than in Group I [ 39 weeks versus 40 weeks, p = 0.009]. At the same time, no statistically significant differences were found between the groups in the modes of delivery. In particular, cesarean section was performed in 67 (19%) pregnant women in Group I, and in 24 (20%) patients in Group II (p = 0.98). Natural delivery was recorded in 282 (80%) pregnant women in Group I and in 97 (80%) patients in Group II (p = 0.98). No statistically significant differences were found between the groups on the indicators of baby weight at birth (p = 0.69).

However, the most important results were obtained when assessing the time of insulin therapy initiation after the detection of GDM. According to clinical guidelines, if two or more non-target glycemic indicators were recorded while following dietary and physical activity recommendations, i.e. the planned glycemic indicators were not achieved within 1–2 weeks of self-monitoring, then pregnant women with GDM were recommended to proceed with insulin therapy to achieve carbohydrate metabolism compensation. The results of the analysis conducted in Group II revealed that 37 (30.6%) patients were prescribed insulin at the first week of the diet therapy, while only 25 (20.7%) patients initiated the insulin therapy after 2 weeks of ineffective treatment against the background of diet. Besides, in 59 (48.8%) pregnant women with GDM, insulin was added after more than 2 weeks of unsuccessful diet therapy. On average, insulin therapy was prescribed at 20 ± 11 weeks of pregnancy.

The insulin administration regimens in patients of Group II require a more detailed description. More specifically, 15 pregnant women (12.4%) received only basal insulin, 15 patients (12.4%) took only bolus insulin, while 91 pregnant women (75.2%) used the basal-bolus scheme. Moreover, when analyzing the average insulin doses, it turned out that in the 15 patients (12.4%) who received only basal insulin, the average dose of prolonged insulin was 4.7 U, in 15 pregnant women (12.4%) who used only basal insulin, the average dose of ultra-short insulin was 5.8 U, and in 91 patients (75.2%) who used the basal-bolus insulin therapy scheme, the average dose of ultra-short insulin was 5.9 U, the average dose of prolonged insulin was 8.3 U.

Discussion

The obtained data make it possible to assert that the clinical characteristics of patients who received alternative schemes of GDM treatment did not differ significantly. For instance, the objective data of pregnant women with GDM, such as age, blood pressure, weight, and BMI, did not differ significantly in each trimester and before delivery. The biochemical blood tests revealed that patients receiving insulin therapy had higher levels of urea (p = 0.009), AST (p = 0.03), and ALT (p = 0.02), as well as higher levels of prothrombin (p = 0.02) and D-dimer (p = 0.047) according to the results of the coagulogram. However, these values did not go beyond the reference values of the norm laboratory tests. So, it made it possible to assert that, in general, there were no clinically significant differences in these patients.

In addition, ultrasound examinations did not provide any useful information. Although patients who followed diet therapy had a higher fetal weight in the first trimester (p = 0.03), subsequent fetal weight indicators up to delivery were comparable in both groups. Analysis of the pharmacological history showed that patients who received insulin therapy also took progesterone preparations significantly more often [ 61 (17%) patients versus 32 (26%) ones, p = 0.02], while the use of other drugs did not differ among pregnant women in the studied groups, which also indicated the absence of clinically significant differences between the patients.

The obtained data on the concomitant diseases, such as overweight and obesity, anemic syndrome, hypothyroidism syndrome, and arterial hypertension, as well as on the complications, including preeclampsia and neonatal hypoglycemia, were also comparable between the studied groups of patients.

The results on the term of delivery revealed the earlier term in patients who received insulin therapy [ 39 weeks versus 40 weeks, p = 0.009], which might be associated with the protocols for monitoring pregnant patients and stricter compliance by pregnant women with the recommendations of the obstetrician-gynecologist. However, the types of delivery, namely the ratio of cesarean section and natural childbirth, were also absolutely comparable between the groups of patients.

The absence of significant differences in the clinical characteristics between patients who received alternative schemes of GDM treatment might be related to the unjustified prescription of insulin therapy, when, at a non-recurrent increase in the glucose level against the background of a diet violation, patients were immediately administered small doses of insulin, although it would have been quite possible to continue diet therapy, eliminating nutritional disorders. This fact is also confirmed by the untimely crossover of pregnant women to insulin therapy. In particular, the study demonstrated that insulin was added to the treatment plan for 59 (48.8%) pregnant women with GDM after more than 2 weeks of unsuccessful diet therapy at an average of about 20 weeks of pregnancy. In addition, this assumption was confirmed by the extremely low average doses of administered insulin, which were no more than 6 U for ultra-short and no more than 9 U for prolonged insulin. This also attested that there was no need to transfer these patients to insulin therapy in general, indicating only the need to correct their diet and continue diet therapy.

Conclusion

Currently, researchers continue active investigations of different GDM subtypes, the identification of which may be of extremely important clinical significance. For this purpose, we analyzed the clinical characteristics of patients with GDM including general clinical, phenotypic, anamnestic, biochemical, and ultrasound features, depending on the regimens of achieving targeted glycemic values. Despite the fact that no significant differences were found among the studied groups, the obtained results do not exclude the existence of different GDM subtypes and suggest an unjustified prescription of insulin therapy in cases where a more intensive change in the diet was required. Moreover, the identification of GDM subtypes can determine the choice of tactics and timeliness of patient treatment, which will promote improving the results on the prognosis of the patient’s condition.

Certainly, a further study of this problem is required to determine, first of all, the criteria for diagnosing different GDM subtypes depending on the mechanisms underlying the pathogenesis of carbohydrate metabolism disorders. The data obtained could be used to develop a new personalized approach to managing patients with GDM, depending on the specific subtype of GDM in addition to existing algorithms. This will significantly reduce the risks of adverse pregnancy outcomes and further glucose intolerance in the postpartum period.

Limitations of the study. This study has several limitations. The work was conducted only in several large medical institutions in Rostov-on-Don, which have sufficient experience in managing patients with GDM, but did not cover all medical organizations in the city involved in monitoring pregnant women. Another limitation is the number of patients involved in the study. In addition, the study only conducted a retrospective analysis of medical records without the ability to clarify missing information directly from patients, as well as determine the GDM subtypes on the basis of insulin resistance indices.