The incidence of GDM is rising worldwide; it is now considered to be the most common medical complication during pregnancy. For example, in the United States, the GDM prevalence range has increased from 6.9% in 2019 to 8.0% in 2023 and in Europe from 5.4% in 2016 to 7.8% in 2021 (Anastasiou et al., 2020; Wang et al., 2022). This is happening for multiple reasons (described below), and screening programs contribute to increasing cumulative prevalence, no matter what diagnostic criteria are used (Reece et al., 2009).

The diagnostic criteria for GDM vary and remain controversial, complicating the comparison of research data. In 2017 in most countries, there has been a move towards the more strict diagnostic criteria advocated by the International Association of the Diabetes and Pregnancy Study Groups (IADPSG)/WHO (Metzger et al., 2010; Saeedi et al., 2021) and this has resulted in a general increase in the overall prevalence of GDM (Song et al., 2018). According to these recommendations, an oGTT is performed by measuring the plasma glucose concentration while fasting and then one and 2 hours after ingesting 75-g of glucose. But, for example, in the United States, GDM is most commonly diagnosed using a two-step approach, with a non-fasting 50-g glucose challenge test followed, only if abnormal, by a 100 g 3-h oGTT, where at least two abnormal values are required; the one-step 75-g oGTT using IADPSG criteria (≥1 abnormal value) is used less often (Committee ADAPP, 2024). In Australia and New Zealand, new, higher cut-offs for positive oGTTs are recommended from 2025 onwards, with the aim of reducing overdiagnosis in women at lower risk and focusing specifically on women with a higher likelihood of complications (Sweeting et al., 2025). Table 3 shows recommendations and GDM criteria used in different parts of the world.

Higher BMI before pregnancy, older maternal age, positive family history, previous occurrence of GDM and history of large for gestational age babies are all considered as crucial risk factors for developing GDM. Studies further suggest the contribution of additional factors, including polycystic ovary syndrome, multigravidity, increased weight gain during pregnancy, higher glycaemic variability, a sedentary lifestyle, smoking and pre-existing hypertension (Plows et al., 2018; Ustianowski et al., 2023). Multiple studies have focused on non-modifiable factors including genetics. Several associations were ascertained between GDM and variants in genes for insulin receptor, glucokinase, melatonin receptor, potassium channel, hepatocyte nuclear factor, peroxisome proliferator receptor, etc. Also, fat mass and obesity associated genes were studied in this context. Most of those genes were associated with T2DM too, which underlies the hypothesis of a shared genetic basis for GDM and T2DM (Kwak et al., 2012).

Metabolomics has been used in pregnancy to assess metabolic profiles by measuring numerous low-molecular-weight metabolites. However, studies on amino acids, free fatty acids and conventional metabolites show inconsistent results in GDM, although women with higher insulin resistance and hyperglycaemia exhibit metabolic changes, similar to non-pregnant insulin-resistant individuals (Alesi et al., 2021).

Maternal diet is an important non-genetic risk factor for GDM(Tobias et al., 2012) not only in quantitative terms (influencing body weight and composition), but also qualitative differences in food intake may contribute to the development of GDM and thus constitute a modifiable environmental risk factor (Bartáková et al., 2018). Diet is also a powerful modulator of the gut microbiota, whose impact on insulin resistance and the inflammatory response in the host is well known. Changes in the gut microbiota composition in the sense of reduced diversity have been described in pregnancies either before the onset of GDM or after its diagnosis (Koren et al., 2012). It remains unclear if the changes in the gut microbiota contribute to or are a consequence of the development of GDM. Finally, the microbiota of GDM patients can be transmitted to the offspring and colonisation before birth by specific taxa associated with GDM occurs (Wang et al., 2018).

Whether there are other sensitive markers that could be identified ans exploited for diagnostic purposes using more complex individual-level data such as omics, and if these can feasibly be implemented in clinical practice remains unknown and will be important to consider in future studies (Benham et al., 2023).

Lifestyle modification recommendations and diabetic diet are the first-line strategies in the management of GDM(49). These are very similar to those recommended for patients with T2DM, but regarding the incentive of pregnancy and foetal nutrition requirements, dietary arrangements are not so strict. The emphasis is on optimising diet composition and limiting foods with a high glycaemic index, rather than on strict carbohydrate or caloric restriction. Similarly, physical activity must be of a reasonable intensity and character.

Regular self-monitoring of glucose levels is a very important integral part of the treatment strategy for all patients with diabetes, including women with GDM. Typically, glucose levels are measured in capillary blood using a blood glucose meter. Nowadays, continuous glucose monitoring (CGM), which provides patients with real-time feedback on glucose levels from interstitial fluid, has become a standard in developed countries. It significantly improves adherence among patients with diabetes, which correlates with better health outcomes (Battarbee et al., 2024). A randomised controlled trial found that the use of CGM reduced the incidence of baby macrosomia (Yu et al., 2014), but another one found no significant difference in the large infant risk between CGM and self-monitored glucose in women with gestational and pre-pregnancy diabetes (Voormolen et al., 2018). So, for now, there is limited data on the effectiveness of CGM for improving pregnancy outcomes in women with GDM, but experience in T1DM patients suggests that CGM could also be beneficial for other types of diabetes, including GDM. The limitation of CGM’s wide application is its higher cost, as well as the potential for users to encounter difficulties interpreting CGM data, underscoring the need for individualised care approaches.

Pharmacotherapy of diabetes in pregnancy is, in general, limited by the lack of safety data for the majority of oral antidiabetic drugs. There is also variability in clinical practice regarding the choice of first-line drug and the initiation of GDM treatment when diet and lifestyle changes are insufficient. Studies show that maternal characteristics, including body mass index (BMI)≥30 kg/m2, family history of T2DM, prior history of GDM and higher glycated haemoglobin (HbA1c), increased the likelihood of the need for insulin treatment (Alvarez-Silvares et al., 2022). The selected insulin or insulin analogue, along with its dosing frequency and dosage, is tailored to the measured glucose profile. Over the last years, metformin has been increasingly used for GDM treatment (as a substitute for insulin) and has shown comparable or even better results in maternal or perinatal outcomes (Picón-César et al., 2021; Spaulonci et al., 2013). Since metformin crosses the placenta, its effects on fetal and childhood growth are uncertain. Some studies suggest no difference in growth and developmental outcomes between children of mothers treated with metformin versus insulin; on the other hand, a systematic review found that metformin-exposed foetuses showed accelerated postnatal growth, leading to higher childhood BMI, which is linked to adverse cardiometabolic outcomes (Picón-César et al., 2021). There is, therefore, a strong need for further research on the long-term effects of intrauterine exposure to metformin.

Additional drugs with a potential for use in pregnancy are glyburide/glibenclamide, which belongs to the group of sulfonylurea-based antidiabetic drugs and has a low rate of crossing the placenta. The few studies conducted in GDM do not yet fully agree on its non-inferiority to insulin and metformin (Hebert et al., 2009; Balsells et al., 2015; Sénat et al., 2018). Recently, inositol, which transfers glucose into cells for conversion into fatty acids, has been investigated. If used early in pregnancy, it can reduce the likelihood of gestational diabetes in at-risk pregnant women by improving insulin sensitivity and glycaemic control (Fabio et al., 2023). However, there is no evidence to date that it can be used directly to treat GDM. Vitamin D deficiency in pregnant women is common throughout the world and the detection rate of GDM is positively correlated with the degree of vitamin D deficiency (Holick, 2007). It has been confirmed that additional vitamin D supplementation can improve hyperglycaemia and delay, or even prevent, the progression of diabetes. Also, a clinical trial conducted in women with GDM who received vitamin D supplementation showed a tendency toward improved glycemic control, but results were not statistically significant. While the preliminary findings are promising, further research is needed to establish definitive conclusions. Also, the modulation of the gut microbiota by dietary interventions during pregnancy is an emerging area of interest, given the potential effects on maternal and, consequently, neonatal health (Ponzo et al., 2019; Gohir et al., 2015). Probiotics are a relatively new intervention; they target mothers’ metabolism, and can reduce blood sugar levels, enhance insulin sensitivity, prevent gestational diabetes and reduce the maternal and foetal complications resulting from it (Benhalima et al., 2019). Using probiotics to modulate the gut microbiota may become part of a comprehensive treatment (de Albuquerque et al., 2024).

GDM is associated not only with maternal but also with neonatal adverse outcomes and represents a significant risk for the offspring. During the perinatal period, GDM increased the incidence of macrosomia, congenital malformations, premature delivery, respiratory distress syndrome during delivery, neonatal hyperbilirubinemia or hypoglycaemia and low Apgar score, both in the first and fifth minutes. Moreover, in childhood, slow psychomotor development was observed, with delays affecting speech, social reactions and motor skills (Kowalczyk et al., 2002; Rizzo et al., 1995). GDM is suspected to impacts offspring neurodevelopmental and cognitive outcomes. An increased risk of deficit/hyperactivity disorder (ADHD) was detected in a meta-analysis (Zhao et al., 2019).

There is increasing epidemiological evidence linking the early-life environmental exposures (i.e., maternal malnutrition/overnutrition, environmental chemicals, stress) and corresponding pathophysiological changes (circulating mediators, neuro-hormonal changes, low-grade inflammation, etc.) with later-life health outcomes–conceptualised as the ‘developmental origins of health and disease’ (DOHaD) paradigm. Compelling studies from animal models have provided strong evidence in support of the DOHaD concept. These have, for example, shown that in utero exposure to maternal diabetes and/or obesity disrupts the development and function of the hypothalamus, predisposing offspring to obesity (Chu and Godfrey, 2020). The actual mediators of DOHaD are numerous and are being explored. For instance, long-term effects on offspring may be mediated by epigenetic changes, i.e., through the regulation of gene activity without changes in the DNA sequence. Epigenetic changes occur, for example, in the form of histone modifications, DNA methylation, or the disruption of the function of non-coding RNAs, including microRNAs, with DNA methylation being the most studied mechanism to date (Dłuski et al., 2021). In pregnant women with GDM, epigenetic studies have shown altered gene methylation and chromatin modifications in the baby’s DNA, providing a potential substrate for disrupted gene regulation (Salbaum and Kappen, 2012). However, further follow-up will be necessary to determine the long-term effects of GDM on offspring of women diagnosed with GDM during pregnancy.

Offspring of GDM mothers are, in any case, more susceptible to suffering from cardiovascular diseases (CVD) and childhood obesity or to developing T2DM more frequently in subsequent life (Metzger et al., 2010; Metzger et al., 2008). Some studies also show a higher risk of allergic diseases or other immunodeficient diseases (Chu and Godfrey, 2020). For an overview, see Table 4.

Women with GDM have an increased risk of not only adverse perinatal outcomes (i.e., during pregnancy up to 1 year postpartum) such as hypertension or preeclampsia, complications during delivery, difficult breastfeeding, etc., but also of persistence of abnormalities in glucose metabolism postpartum. They confer a 7× higher likelihood of developing any glucose metabolism abnormality (especially prediabetes or T2DM in the future; T2DM is expected to develop in 20%–50% of these women within 10–20 years (Bellamy et al., 2009; Bartáková et al., 2015). Several studies aimed at the persistence of abnormalities in glucose metabolism postpartum, respectively early after GDM pregnancy, briefly, 15%–20% of women with GDM complicated pregnancy had T2DM or impaired glucose tolerance (IGT) (Benhalima et al., 2019). Studies focusing on other forms of diabetes developing after a GDM pregnancy are rare. For example, our study with 305 GDM women showed 16.7% prevalence of any form of glucose metabolism abnormality; of those, 62.7% had prediabetes, 37.3% had manifest diabetes (31.4% were diagnosed as T2DM, 5.9% as T1DM) (Bartáková et al., 2015); another study has shown the prevalence of T1DM after a GDM pregnancy 7% (Füchtenbusch et al., 1997).

The risk of a persisting abnormality of glucose metabolism after delivery is not the same for every woman with GDM (White et al., 2020). Factors such as fasting glucose levels, BMI, HbA1c, the total number of abnormal glucose values during antenatal oGTT, and the timing of diagnosis (early or late GDM) are important in predicting postnatal glucose intolerance (Gupta et al., 2024). Some studies indicate a higher fasting plasma glucose during pregnancy oGTT test and lower BMI in those GDM women, who later developed T1DM (Unnikrishnan et al., 2016) and, opposite, higher BMI and higher levels of glucose in 120th min of oGTT in pregnancy in those with T2DM development in their later life (Cypryk et al., 2005). The risk could also reflect likely etiopathogenetic heterogeneity of the disease (e.g., various contributions of genetic, environmental and other factors). On the contrary, breastfeeding has been shown to significantly lower the risk of developing postnatal diabetes both in children and women with a history of GDM (Taylor et al., 2005). So far, it seems that the longer and more intensive exclusive breastfeeding, the greater the reduction of the risk of diabetes outbreaks (Eades et al., 2024).

Predicting postnatal diabetes after GDM is now feasible through various risk prediction models, which utilise a combination of clinical data, glucose test results and medical history. Many of which, according to their authors, show relatively good discriminatory power (AUC 0.725–0.940). The variables that appear most frequently in the models are age, BMI, family history, physical activity levels, ethnicity, insulin requirement in the therapy, etc., whereas the most significant appear to be blood sugar values in oGTT and HbA1c (Köhler et al., 2016; Parkhi et al., 2023; Lee et al., 2007). The addition of lipid markers could significantly improve the reclassification of patients as well as early postpartum data (e.g., glycemia, weight) (Lappas et al., 2015; Belsti et al., 2024). Interestingly, polygenic risk scores for T2DM are associated with an incidence of T2DM after GDM, but they provide only a modest improvement in discrimination and risk reclassification, suggesting limited incremental value for routine clinical risk stratification at present (Choi et al., 2024). However, most models suffer from significant methodological limitations, namely, insufficient description of predictor selection, limited internal as well as external validation, leading to a generally higher risk of bias, etc. So, challenges remain in standardising these tools, and there is a need for further research and external validation to make them applicable for widespread clinical use (Belsti et al., 2024). Further examination of the metabolome and its incorporation into predictive models, together with the routine evaluation of additional biomarkers in the first trimester of pregnancy - such as antibodies used for T1DM diagnosis or the measurement of leptin, adiponectin, and resistin (Kapustin et al., 2020) - might improve the prediction of future disorders of glucose metabolism; however, their clinical significance is yet to be clearly established (Atarod et al., 2020; Sun et al., 2024; Huhtala et al., 2023; Liu et al., 2024; Semnani-Azad et al., 2024).

By identifying women with GDM who are at higher risk of persisting some form of abnormality of glucose metabolism early postpartum, we can implement earlier intervention and ensure more intense postpartum follow-up. Effective lifestyle modifications, especially dietary changes, physical activity, maintaining a normal BMI, and breastfeeding, can significantly lower the chances of developing DM (Ikoh Rph and Tang Tinong, 2023). So, there is a strong need for repeated check-ups and personalised education, including targeted interventions (Elomrani et al., 2025). Unfortunately, the participation of GDM women in postpartum follow-up programs is still unsatisfactory–more than half of patients did not pass a recommended OGTT test after delivery (Ferrara et al., 2009; Keely, 2012), and moreover, longer follow-up studies for GDM women and their children are rare.