ΔΙΑΒΗΤΗΣ ΚΥΗΣΗΣ

Gestational Diabetes

Xiao-Yu Wang, MD

Fellow, Department of Obstetrics and Gynecology, Division of Maternal Fetal Medicine, The Ohio State University Wexner Medical Center, Columbus, OH.

Erin M. Cleary, MD

Assistant Professor, Department of Obstetrics and Gynecology, Division of Maternal-Fetal Medicine. Indiana University School of Medicine, Indianapolis, IN.

Stephen F. Thung, MD

MSCI Professor, Department of Obstetrics and Gynecology, Division of Maternal Fetal Medicine, Yale University, New Haven, CT.

Kartik K Venkatesh, MD, PhD

Associate Professor, Department of Obstetrics and Gynecology, Division of Maternal Fetal Medicine, The Ohio State University Wexner Medical Center, Columbus, OH.

Elizabeth O. Buschur, MD, FACE

Associate Professor, Division of Internal Medicine, Department of Endocrinology, Metabolism, and Diabetes, The Ohio State University Wexner Medical Center, Columbus, OH.

Corresponding author.

Last Update: August 28, 2024.

ABSTRACT

Gestational Diabetes (GDM) is characterized by glucose intolerance first manifesting during pregnancy and is an important risk factor for abnormal birthweight including large-for-gestational age birth, birth injury, and neonatal metabolic alterations–particularly when persistent maternal hyperglycemia exists. Individuals with GDM face short-term pregnancy complications, such as cesarean section and hypertensive disorders, and long-term complications, including an increased lifetime risk for type 2 diabetes and cardiovascular disease. Many of these complications associated with GDM can be moderated with lifestyle changes and pharmacotherapeutic interventions to reduce maternal hyperglycemia and thereby improve maternal and neonatal health. For complete coverage of all related areas of Endocrinology, please visit our on-line FREE web-text, WWW.ENDOTEXT.ORG.

PREVALENCE AND PATHOPHYSIOLOGY

Gestational diabetes (GDM) is a major problem in the US with a rapidly rising prevalence ranging from 6 to 20% of pregnancies across the world (1–4). The prevalence is highest in racial and ethnic groups that have a higher incidence of type 2 diabetes mellitus (T2DM): Hispanic, Black, American Indian, and Asian or Pacific Islander individuals (1). In the US, the prevalence of GDM was 7.8 per 100 births in 2020 with a higher annual percent change from 5% annually from 2016-2019 to 13% in 2020 (5). There are significant differences in the rate of GDM also by rural/urban status, age, body mass index (BMI), and state of residence (5,6).

Insulin resistance is paired with increased insulin secretion during pregnancy; however, among individuals with GDM, inadequate β-cell compensation relative to the level of insulin resistance results in failure to maintain euglycemia (7). GDM is caused by carbohydrate intolerance due to abnormalities in at least 3 aspects of fuel metabolism: insulin resistance, impaired insulin secretion, and increased hepatic glucose production (8,9).

Although insulin resistance is a universal finding in pregnancies with GDM, the cellular mechanisms for insulin resistance are multifactorial. Insulin binding to its receptor is unchanged in pregnant individuals (10). Pregnancy reduces the capacity for insulin-stimulated glucose transport independent of obesity, due in part to a tissue-specific decrease in insulin receptor phosphorylation and decreased expression of Insulin Receptor Substrate-1 (IRS-1), a major docking protein in skeletal muscle (11). In addition to these mechanisms, in muscle specimens from individuals with GDM, IRS-1 is further decreased and there are reciprocal and inverse changes in the degree of serine and tyrosine phosphorylation of the insulin receptor (IR) and IRS-1, further inhibiting insulin signaling (12). Individuals with GDM also have higher circulating free fatty acids (FFA) and reduced peroxisome proliferator-activated receptor (PPAR) expression in adipose tissue, a target for thiazolidinediones (13). There is evidence for a decrease in the number of glucose transporters (GLUT-4) in adipocytes in individuals with GDM and an abnormal translocation of these transporters that results in the reduced ability of insulin to recruit them to the cell surface, which contributes to the overall insulin resistance with GDM (14). In individuals with GDM, serum adiponectin levels decrease and leptin, IL-6, and TNFα increase (15).

Although dysglycemia usually remits after pregnancy, individuals with a history of GDM have a nearly 10-fold higher risk of progressing to type 2 diabetes than those without GDM and up to 70% of individuals diagnosed with GDM will develop T2DM later in life (16–18). Both GDM and T2DM are further exacerbated by increasing obesity and age. Thus, pregnancy is a “stress test” for the eventual development of glucose intolerance outside of pregnancy, and GDM may represent an unmasking of the predisposition of T2DM induced by the hormonal changes of pregnancy (19,20).

DATA TO SUPPORT THE SCREENING, DIAGNOSIS, AND TREATMENT OF GESTATIONAL DIABETES

As early as the 1940s, glucose intolerance and GDM distinct from pregestational DM were recognized to have adverse maternal and perinatal outcomes (21). Over the course of the following eight decades, candidates for screening or testing as well as the diagnostic criteria of GDM were debated (22,23). Early on, value was placed on recognition of GDM to identify individuals at risk for T2DM (24). More recently, robust studies have demonstrated the more immediate obstetric and perinatal benefits of screening, diagnosing, and treating GDM (1,24,25).

The first was a landmark trial conducted in Austria and New Zealand referred to as the ACHOIS trial (Australian Carbohydrate Intolerance Study in Pregnant Women), which demonstrated reduced serious perinatal outcomes with intervention/treatment of GDM versus no intervention (1% versus 4%) (24). This RCT enrolled 1000 pregnant individuals with either ≥1 risk factor for GDM or positive 1-hour 50 gm glucose challenge test (GCT) (≥140mg/dL) after completion of a blinded 75 gm glucose tolerance test (GTT) at 24-34 weeks gestation and demonstrated no severe glucose impairment. Individuals were randomized to receive dietary advice, self-monitoring of blood glucose (SMBG), and insulin therapy as needed to achieve fasting glucose <99 mg/dL and 2-hour postprandial values <126 mg/dL versus routine care. The primary outcome of fetal or neonatal death, shoulder dystocia, bone fracture, and nerve palsy were reduced in the intervention group compared with routine care (1% versus 4%). Although the induction of labor rate was higher in the intervention group, the cesarean delivery rate was not different.

A second landmark randomized controlled trial (RCT), the National Institute of Child Health and Human Development Maternal- Fetal Medicine Units Network study (NICHD MFMU Network), demonstrated significant differences in meaningful secondary outcomes (mean birthweight, neonatal fat mass, large for gestational age infants, birthweight >4000 g, shoulder dystocia, cesarean delivery, and hypertensive disorders of pregnancy) by treating mild GDM (25). A total of 958 pregnant individuals who met criteria for mild GDM between 24-31 weeks were randomly assigned to usual prenatal care (control) or dietary interventions, SMBG, and insulin therapy if necessary (treatment group). Although there was no significant difference in groups in the frequency of the composite outcome and no perinatal deaths in this population with very mild GDM, there were significant reductions with treatment in several pre-specified secondary outcomes. Furthermore, treatment of mild GDM was also associated with reduced rates for hypertensive disorders of pregnancy.

There is also compelling data that the risk of adverse maternal and fetal outcomes from maternal carbohydrate intolerance is along a graded continuum (26,27). The Hyperglycemia and Adverse Outcomes (HAPO) trial enrolled 25,505 pregnant individuals at 15 centers in nine countries, with participants completing a 2-hour 75 gm GTT at 24-32 weeks’ gestation (27). Data remained blinded if the fasting glucose ≤105 mg/dL and the 2-hour plasma glucose was ≤200 mg/dL. This trial demonstrated that a fasting blood glucose (FBG) ≥92 mg/dl, a 1-hour value ≥180 mg/dl, or a 2-hour value of ≥153 mg/dl increased the risk by 1.75-fold for large for gestational age (LGA or more than 90^th percentile for weight of all babies of the same gestational age) and an elevated cord-blood C-peptide consistent with fetal hyperinsulinemia. Furthermore, the fasting glucose was more strongly predictive of these outcomes than the 1-hour or 2-hour value stressing the importance of fasting glucose levels in predicting poor perinatal outcomes. The results also indicated a strong and continuous association with these outcomes and maternal glucose levels below those considered diagnostic of GDM.

DIAGNOSIS OF GESTATIONAL DIABETES

The implications of diabetes recognized for the first time in pregnancy on both maternal and perinatal outcomes have been known for nearly a century (21,28). Nevertheless, substantial controversy persists when considering which individuals warrant screening or outright diagnostic testing, at which gestational age this should occur, and the laboratory cutoffs which should confirm the diagnosis and prompt possible intervention.

Evaluation for what we currently define as GDM, for “early GDM” or for T2DM first diagnosed in pregnancy may take place in a risk-based or universal manner. In general, the basic tenants of screening as defined by the World Health Organization (WHO) are met: GDM and DM are significant health problems with significant consequences if left untreated, a suitable test exists, with benefits of screening outweighing the risks (29). The lack of consensus generally results from concerns about cost-effectiveness of differing strategies and psychological and public health burdens with increased prevalence of the condition.

Unfortunately, the historical definition of GDM as a glucose-intolerant state with onset or first recognition during pregnancy allows for inclusion of both unrecognized, pregestational, overt diabetes in addition to “true” gestational diabetes resulting from the physiologic and hormonal changes of pregnancy. Since individuals with undiagnosed pregestational diabetes are at increased risk for both maternal and fetal complications, including major malformations if their hemoglobin A1c is ≥ 6.5% in the first trimester, the International Association of the Diabetes and Pregnancy Study Groups (IADPSG) recommends that GDM be only diagnosed if the glucose intolerance was identified in pregnancy and the pregnant individual did not qualify for pre-existing (overt) diabetes (30). The details and nuanced considerations of the current understanding of GDM as well as screening and testing modalities are described in the literature (31–33).

In the US, the US Preventive Services Task Force (USPSTF) and the American College of Obstetricians and Gynecologists (ACOG) recommend universal screening for all pregnant individuals at 24-28 weeks gestation since prior use of historic factors alone failed to identify 50% of patients with GDM (1). ACOG further supports a two-step process involving a 50 gm,1-hour glucose challenge test (GCT) followed by a 100 gm, 3-hour oral glucose tolerance test (GTT) (1). The diagnosis of GDM is made if two or more abnormal values are seen in the 3-hour GTT. Although elevation of just one out of four values in the 3-hour GTT is still associated with adverse outcomes, there is not clear evidence that this subset of patients would benefit from treatment (1,34). The one-step International Association of the Diabetes and Pregnancy Study (IADPSG) approach is utilized more frequently internationally and is supported by organizations such as National Institute for Health and Care Excellence (NICE) in UK and Royal Australian and New Zealand College of Obstetricians and Gynaecologists (RANZOG) (35,36). The American Diabetes Association (ADA) continues to recognize that there is no clear evidence which supports IADPSG versus traditional ACOG two-step screening approach (37).

Table 1.

Gestational Diabetes. Screening and Diagnostic Criteria, performed at 24-28 weeks’ gestation for individuals without evidence of pregestational

diabetes^α

α: Note that while most international and national guidelines have moved toward universal testing, the United Kingdom’s National Institute for Health and Care Excellence continues to recommend risk-based GDM testing (35).

β: Cutoff of 140mg/dL results in sensitivity of 70–88% and specificity of 69–89% and requires GTT in ~15% patients; cutoff of 130mg/dL results in sensitivity of 88–99% and specificity of 66–77% specific and requires GTT in ~25% patients.

¥

: ACOG allows for consideration of diagnosis with ≥1 criteria met.

The controversy with the diagnosis of GDM relies heavily on the outcomes studied. The IADPSG recommendations are based on the HAPO trial, which showed that a single value on a 75 gm 2-hour GTT resulted in a 1.75-fold increased risk of LGA and thus should be the basis for the diagnosis; however, critics disagree (38,39). Critics contend that a 2.0-fold increase in LGA risk instead of 1.75 could have been chosen which would not have appreciably increased the prevalence of GDM over the ACOG criteria (40,41). Nearly 90% of all individuals who met criteria for GDM using the 75 gm 2-hour GTT were diagnosed based on the fasting blood glucose (FBG) and 1-hour values, raising the question of whether the 2-hour value is worth the extra time and cost (42). Additionally, evaluation of higher glycemic criteria (fasting glucose ≥99 and 2-hour level glucose ≥162 mg/dl) for the diagnosis of gestational diabetes by the Gestational Diabetes Mellitus Trial of Diagnostic Detection Thresholds (GEMS) trial group did not increase the risk for LGA infant (43).

Currently there is no consensus about the adoption of the IADPSG criteria over the ACOG criteria. The NIH held a Consensus Conference in March of 2013 (44). They acknowledged that the HAPO study was the first to demonstrate that glycemic thresholds currently lower than the ACOG diagnostic criteria thresholds correlated with LGA and adopting the 75 gm GTT would be beneficial in standardizing diagnostic criteria internationally. However, they concluded that there were insufficient data from RCTs demonstrating that adopting the lower glucose thresholds would significantly benefit the much larger population of women who make the diagnostic criteria for GDM based on the IADPSG criteria and such adoption could markedly increase cost of treatment. Further, there was a concern that adopting the IADPSG criteria could triple the prevalence of GDM, potentially outstripping the resources to treat it. Recent clinical trials comparing screening strategies confirm that the one-step approach has been associated with an upwards of threefold increase in the prevalence of GDM compared to the two-step approach, but there is no clear evidence that the increased diagnosis is associated with improved maternal or perinatal outcomes (37,39,45,46).

At the consensus conference, it was also argued that it is not clear how much the increased risk of LGA at lower glucose thresholds observed in the HAPO trial on which it was based was due to maternal obesity or mild hyperglycemia. A retrospective review of nearly 10,000 individuals who were diagnosed with GDM using the IADPSG criteria showed an overall GDM prevalence of 24%. After excluding individuals who required treatment for GDM, 75% of GDM individuals were overweight or obese. Although GDM nearly doubled the risk of LGA over obesity alone (22.3% versus 12.7% respectively), in individuals without GDM, 21.6% of LGA was attributable to being overweight and obese. The combination of GDM in addition to being overweight or obese did not add much to the attributable risk for LGA and accounted for 23.3% of LGA infants (47).

The threshold levels for the 3-hour GTT have been debated. The initial diagnostic thresholds for the 3-hour GTT were initially established by O’Sullivan and Mahan (48). National Diabetes Data Group and Carpenter and Coustan criteria have subsequently been developed since with Carpenter and Coustan criteria suggesting more stringent thresholds (49,50). Further retrospective investigations have shown that the lower Carpenter and Coustan criteria may diagnose more patients that could benefit from treatment of GDM (51–53). Therefore, ACOG and ADA recommend the Carpenter and Coustan criteria (37,54).

Recently, Hillier et al published the results of their pragmatic, randomized trial comparing one-step screening with two-step screening in which nearly 24,000 individuals were randomized (45). The diagnosis of GDM was roughly doubled using the one-step approach versus the two-step approach, at rates of 16.5% and 8.5% respectively (unadjusted relative risk, 1.94; 97.5% confidence interval [CI], 1.79 to 2.11) (40). Importantly, there were no significant differences between diagnostic approaches with the primary outcomes (LGA infants, perinatal composite, gestational hypertension or preeclampsia, and primary cesarean delivery). Potential limitations include a sample size underpowered to detect differences in the groups, treatment of individuals with a single elevated GTT value, and suboptimal adherence to protocols (38). A meta-analysis comparing one-step and two-step screening methods showed a twofold risk of the diagnosis of GDM with the one-step approach without a difference in the risk of LGA infants (39).

Screening Prior to 24 Weeks Gestation- Who, Why, and How?

Several factors have contributed to newer recommendations to consider early screening for diabetes: the rising incidence of T2DM because of the obesity epidemic, in addition to a better understanding of the risks associated with hyperglycemia early in pregnancy, such as congenital anomalies and miscarriage (55). In an obstetric setting, the best time to screen for and diagnose T2DM is as early as possible in the pregnancy, ideally in the first trimester before the placental effects causing insulin resistance have begun. There is not a recognized lower gestational age cutoff at which insulin resistance can be attributed to placental effects, though this likely occurs prior to the GDM screening range of 24-28 weeks.

There is conflicting evidence that screening for GDM prior to 24 weeks gestational age is beneficial. A RCT of 962 individuals examining early screening for GDM in individuals living with obesity compared to standard of care did not show reductions in composite perinatal outcomes (56). Even if an individual is diagnosed with GDM prior to 20 weeks, only modest benefit in composite perinatal outcomes was seen with early treatment (57). As a result, early screening for GDM is not routinely recommended, but screening high-risk individuals for overt diabetes is recommended. The 2024 ADA guidelines recommend screening individuals at the first prenatal visit if BMI >25 kg/m2 (or >23 kg/m2 in Asian Americans) and one or more risk factors (Table 2). Universal screening of pregnant individuals <15 weeks of gestation for undiagnosed pregestational diabetes could also be considered especially in populations with a high prevalence of undiagnosed diabetes. The IADPSG/ADA recommends that individuals diagnosed for the first time in pregnancy prior to 24 weeks should be considered as having overt diabetes following the same criteria for diabetes outside of pregnancy (Table 3) (37,42). ACOG does not recommend screening for GDM prior to 24 weeks, but in line with ADA recommendations it does recommend screening for overt diabetes early in pregnancy (54). The USPSTF concludes there is insufficient evidence to recommend screening for GDM before 24 weeks gestation (58).

The method of screening for diabetes in early pregnancy is also controversial. The ADA recommends screening for abnormal glucose metabolism in early pregnancy may be accomplished with fasting glucose <110 mg/dL or A1c <5.9% (37). ACOG recommends following the same diagnostic criteria for diabetes outside of pregnancy including A1c value ≥6.5% or higher, fasting plasma glucose value ≥126 mg/dL, or 2-hour plasma glucose value ≥200 mg/dL during a 75-g GTT, or random plasma glucose value ≥200 mg/dL in patients with classic hyperglycemia symptoms (54). Individuals with evidence of impaired glucose tolerance or prediabetes such as A1c value 5.7–6.4% or 2-hour glucose value between 140 and 199 mg/dL on the 75-g GTT, ACOG recommends nutrition counseling when resources are available (54). If overt T2DM is not diagnosed on early screening, then routine screening for GDM between 24-28 weeks is still recommended.

Table 2.

Risk Factors for Early Diabetes Screening

Table 3.

Early Screening: Identification of Prediabetes and Diabetes

α: ACOG recommends the following values for prediabetes.

β: ADA Standards of Care 2024 suggests that using a threshold fasting glucose of 110 or HbA1c of 5.9% can identify individuals who are at higher risk of adverse pregnancy and neonatal outcomes.

RISKS TO THE MOTHER AND INFANT WITH GESTATIONAL DIABETES

The pregnancy-associated risks to the individual with GDM are an increased incidence of cesarean delivery (~25%), hypertensive disorders of pregnancy (~20%), and polyhydramnios (~20%) (59–62). The long-term risks are related to recurrent GDM pregnancies and the substantial risk of developing T2DM. Individuals with GDM represent a group with an extremely high risk (~50-70%) of developing T2DM in the subsequent 5-30 years (1,16–18,63,64). Individuals with fasting hyperglycemia, obesity, those belonging to a racial/ethnic group with a high prevalence of T2DM (in particular those who self-identify as Asian and Pacific Islander and Hispanic), or who demonstrate impaired glucose tolerance or fasting glucose at 6 weeks postpartum, have the highest risk of developing T2DM (1,3,63). Counseling about diet, weight loss, and exercise is essential and is likely to improve insulin sensitivity.

Thiazolidinediones, metformin, and lifestyle modifications have all been demonstrated to decrease the risk of developing T2DM in GDM women who have impaired fasting glucose or glucose intolerance postpartum (65–67).

The potential risks to the infant from GDM are similar to those in pregnancies complicated by T1DM or T2DM with suboptimal control in the second half of pregnancy (stillbirth, macrosomia, shoulder dystocia, premature delivery, neonatal hypoglycemia, hyperbilirubinemia, and NICU admission). Notably congenital malformations and miscarriage risks would not be observed with later-onset insulin resistance with GDM (1,62,68,69). Like pregestational DM, the degree of hyperglycemia correlates with perinatal risk. Among diet-controlled GDM pregnancies the risk of stillbirth is not increased compared to the general population (70). The developmental origin of health and disease suggests that there can be an association with maternal health on the future health of the child (71,72). In addition to immediate postnatal risks, infants of individuals with GDM are themselves at increased risk for childhood and adult-onset obesity and diabetes (73).

MEDICAL NUTRITION MANAGEMENT AND EXERCISE

Individuals with GDM should be taught home glucose monitoring to ensure that their glycemic goals are being met throughout the duration of pregnancy. The same goals are utilized in GDM pregnancies as in pregestational DM: a fasting glucose <95mg/dL, 1-hour postprandial <140mg/dL and 2-hour postprandial <120mg/dL (1,37). The best therapy for GDM depends on the severity of the glucose intolerance, the individual’s response to non-pharmacologic or pharmacologic treatment, as well as effects on fetal growth. Diet and exercise alone will maintain the fasting and postprandial blood glucose values within the target range in at least 50% of individual with GDM. A 2018 meta-analysis evaluating diet modifications illustrated improvements in fasting and postprandial glucose values and lower need for medical treatment when compared with controls (74). Furthermore, diet modifications were also associated with lower infant birth weight and less macrosomia (74).

Nevertheless, data are limited regarding the optimal GDM diet to achieve euglycemia and avoid perinatal complications of GDM. The current recommended diet for GDM includes consumption of at least 175 gm of carbohydrate, with complex carbohydrates favored over simple carbohydrates, a minimum of 71 gm of protein, and 28 gm of fiber to provide adequate macronutrients and avoid ketosis (75). There is little evidence to support one dietary approach over another, but common practice is three meals and 2-3 snacks daily to distribute carbohydrate intake and reduce postprandial hyperglycemia. The caloric intake and weight gain recommendations are also consistent with what is recommended in individuals with obesity or T2DM. We recommend multidisciplinary care with a dietician or nurse educator familiar with GDM to individualize a dietary plan.

In 2009, the Institute of Medicine (name changed to National Academy of Medicine in 2015), published recommended gestational weight gain (GWG) based on pre-pregnancy BMI, with less GWG recommended for overweight and obese individuals compared to those with normal BMI (76). Nevertheless, follow up studies demonstrated a large proportion of pregnant individuals worldwide had GWG above (27.8%) and below (39.4%) the IOM guidelines, with highest mean GWG and pre-pregnancy BMI observed in North America (77). Furthermore, a variety of adverse outcomes have been observed in the setting of excess GWG, including LGA and macrosomic infants in the GDM population (78–80). As a result, the question has been raised whether weight gain less than the IOM guidelines in GDM pregnancies could improve outcomes. There are studies suggesting that weight gain less than IOM recommendations for overweight GDM women may decrease insulin requirements, cesarean delivery, and improve pregnancy outcomes without appreciably increasing small for gestational age (SGA) (81–83). Further, a third study suggesting that slight weight loss (mean of 1.4 kg) in overweight GDM women decreased birth weight without increasing SGA (84). Larger meta-analyses confirmed these findings but failed to identify an ideal weight gain range for optimal outcomes (85). These findings have yet to be included in GWG guidelines specific to pregnant individuals living with diabetes.

Since postprandial glucose levels have been strongly associated with the risk of macrosomia it has been suggested that carbohydrate restriction to ~33-40% of total calories may be helpful to blunt the postprandial glucose excursions, in addition to preventing excessive weight gain (86). However, the actual dietary composition that optimizes perinatal outcomes is unknown. There is also growing concern that individuals are substituting fat for carbohydrates which has been associated with adverse fetal programming including oxidative stress as well as an insulin resistant phenotype (87,88). Although a low carbohydrate, higher fat diet has been conventionally recommended to minimize postprandial hyperglycemia, a review of the few randomized controlled trials examining nutritional management in 250 GDM individuals suggested that a diet higher in complex carbohydrate and fiber, low in simple sugar and lower in saturated fat may be effective in blunting postprandial hyperglycemia, preventing worsened insulin resistance, and excess fetal growth (89). A more recent trial challenged the traditional low-carb/higher-fat diet and demonstrated that a diet with higher complex carbohydrates and lower-fat reduced fasting blood glucose and infant adiposity (90). Given these trials, a diet of complex carbohydrates is recommended over simple carbohydrates primarily due to slower digestion time which prevents rapid increases in blood glucose.

The role of exercise in GDM may be even more important than in individuals with preexisting diabetes given exercise in some individuals may lessen the need for medical therapy. This idea is similar to the evidence in non-pregnant individuals with diabetes which supports weight training due to increases in lean muscle and increased tissue sensitivity to insulin. A 2013 review showed that in individuals with GDM, five of seven (~70%) activity-based interventions showed improvement in glycemic control or limiting insulin use (91). In most successful studies (3 times/week), insulin needs decrease by 2-3-fold, and overweight or obese women benefited the most with a longer delay from diagnosis to initiation of insulin therapy. Moderate exercise is well tolerated and has been shown in several trials in individuals with GDM to lower maternal glucose levels (92–94). Using exercise after a meal in the form of a brisk walk may blunt the postprandial glucose excursions sufficiently in some individuals that medical therapy might be avoided.

Establishing a regular routine of modest exercise during pregnancy, per ACOG of 30 minutes of moderate-intensity aerobic activity at least 5 days/week, may also have long lasting benefits for the individual with GDM who clearly has an appreciable risk of developing T2DM in the future (1).

MEDICAL TREATMENT OPTIONS

Once the diagnosis of gestational diabetes is confirmed and glycemic control exceeds target ranges despite dietary education and lifestyle changes, pharmacotherapy must be considered. Prior to the 21st century, insulin was the sole medical option for GDM. After the introduction of oral agents such as glyburide and metformin, initiation of these agents increased, with addition or transition to insulin therapy if glycemic control remained suboptimal. In 2018, ACOG joined the ADA in endorsing insulin as first-line therapy and this was reaffirmed in the 2024 ADA Standards of Care. In contrast, the Society of Maternal-Fetal Medicine (SMFM) released a statement recommending metformin as a reasonable first-choice therapy in addition to insulin (1,75,95). The optimal use of oral agents for the management of GDM is an area of ongoing investigation (96,97).

Although there are few data from randomized controlled trials to determine the optimal therapeutic glycemic targets, the standard of care is that individuals who have fasting blood glucose levels >95 mg/dl, 1-hour postprandial glucose levels >140 mg/dl or 2-hour postprandial glucose levels >120 mg/dl be started on medical therapy. In 5 randomized trials it was demonstrated that if insulin therapy is started in individuals with GDM whose maternal glucose values are at target levels on diet alone but whose fetuses demonstrate excessive growth by an increased abdominal circumference (AC) relative to the biparietal diameter (BPD) i.e. body to head disproportion, the rate of fetal macrosomia can be decreased (81). This fetal based strategy using ultrasound at 29-33 weeks to measure the AC in order dictate the aggressiveness of maternal glycemic control has been recommended by the Fifth International Workshop-Conference on Gestational Diabetes and the IADPSG (30,98,99). Evidence also suggests that higher glycemic targets may have similar outcomes to lower glycemic targets (100). The threshold to initiate pharmacologic therapy by maternal fetal medicine specialists can differ, but most providers recommend initiation of pharmacologic therapy if 30-50% of blood glucose values are elevated (101).

Continuous glucose monitoring (CGM) has also been proposed as an adjunct to traditional glucose monitoring. CGM use has proven beneficial in type 1 diabetes in pregnancy, and its use is recommended by ADA and ACOG (75,102,103). However, there is no clear evidence of its benefit in the setting of GDM. One meta-analysis including 6 trials of 482 individuals with GDM showed that CGM use may achieve lower average blood glucose levels, lower maternal weight gain and infant birth weight, but the studies were limited by overall small sample sizes (104). The specific metrics of CGM data, including pregnancy-specific time-in-range, has not been clearly established and further research should be done before widespread implementation of CGM in GDM.

GDM can often be treated with twice daily injections of intermediate or long-acting insulin (i.e., NPH, glargine, detemir) and mealtime injections of lispro or aspart as necessary for postprandial hyperglycemia. Short acting insulin (i.e., lispro or aspart) is preferred over regular insulin due to time of onset and duration to better control postprandial glycemic excursions. See Endotext chapter “Pregestational Diabetes” for details regarding antepartum, intrapartum, and postpartum insulin dosing regimens (105).

Metformin

One of the largest experiences with metformin in the setting of GDM was with metformin initiated later in pregnancy (106). In this randomized, controlled Metformin in Gestation (MIG) trial, 751 individuals with GDM were randomized to metformin versus insulin. Individuals that did not get adequate glycemic control on metformin received insulin. There was no difference in both groups concerning the primary composite outcome (neonatal hypoglycemia, respiratory distress, need for phototherapy, birth trauma, 5- minute APGAR <7), or premature birth. As such, metformin did not appear to increase any adverse outcomes, although it was associated with a slight increase in preterm birth; however, this did not appear to be clinically relevant.

Importantly, 46% of the individuals in the metformin group required supplemental insulin to achieve adequate glycemic control. This study demonstrated interesting metformin benefits including reduced maternal weight gain, improved patient satisfaction, and reduced incidence of gestational hypertension. In a smaller RCT, Ijas et al demonstrated metformin had a 32% rate of requiring supplemental insulin (107). They also noted individuals needing supplemental insulin added to metformin were more likely to be obese, have higher fasting blood glucose levels, and initiated pharmacotherapy earlier. Spaulonci et al randomized 47 individuals with GDM to metformin or insulin and demonstrated significant metformin benefits including: less gestational weight gain, lower mean glucose levels, and lower rates of neonatal hypoglycemia (108). Overall, meta-analyses have demonstrated largely reassuring outcomes for metformin compared to insulin and glyburide (109–112). Early initiation of metformin after diagnosis of GDM did not show benefit in fasting BGs over standard of care (113).

Metformin should be avoided in individuals with renal insufficiency. It is typically prescribed in divided doses starting with 500 mg once or twice daily for 1 week and then increasing to a maximum dose of 2500 mg daily in divided doses with meals. Common side effects include gastrointestinal complaints (occurring in 2.5-45.7% of pregnant individuals in studies) (109). These randomized trials have shown short-term efficacy and safety of metformin use in pregnancy for GDM treatment.

Until recently, long-term safety data of in-utero metformin exposure has been lacking, though several studies have been published commenting on infant and childhood weight, BMI, cardiovascular health, and neurodevelopmental outcomes. The earliest follow-up studies in metformin-exposed infants suggested they had larger measures of subcutaneous fat when compared to those exposed to insulin (110). Another study in PCOS women comparing metformin to placebo showed that although women randomized to metformin gained less weight during pregnancy, at 1 year postpartum the women who used metformin in pregnancy lost less weight and their infants were heavier than those in the placebo group (112). These fetal and neonatal results are likely because metformin is concentrated in the fetal compartment with umbilical artery and vein levels being up to twice those seen in the maternal serum (114,115). Hypothetically if metformin increases insulin sensitivity in the fetus, it might be possible for excess nutrient flux across the placenta to result in increased fetal adipogenesis. More recent studies also identified slight increased weight in metformin-exposed infants, a relationship that may no longer be present after 4 years of age (116,117). Long-term follow up of BMI is conflicting (112,117). A very large study in New Zealand evaluated outcomes at 4 years of age in nearly 4000 individuals, with no differences in growth and development assessments compared to insulin-exposed children (118). The metabolic and weight differences warrant further investigation since similar patterns of low birth weight followed by accelerated growth are associated with adverse long-term outcomes (119). At least one study has failed to demonstrate such a relationship in this population (120). A study of 211 individuals with GDM randomized to insulin versus metformin during pregnancy found similar developmental outcomes by 2 years of age (111).

In review, the ADA and ACOG note that insulin is the first-line agent for treatment of GDM if lifestyle changes have not achieved glycemic targets (1,75). The ADA notes that although individual RCTs have shown short-term benefits and safety of metformin and glyburide, long- term safety data are lacking (75). Both organizations acknowledge that 20-45% of women fail metformin monotherapy necessitating that insulin be added (1). Counseling is necessary to explain to women that although current data do not demonstrate any adverse short-term outcomes, there are concerns about placental transfer of metformin, potential increased preterm birth, and lack of data on long term outcomes of fetuses exposed to metformin in-utero, metformin’s effect on fetal insulin sensitivity, hepatic gluconeogenesis, and the long-term fetal programming implications are unknown. SMFM suggests that metformin is a reasonable and safe alternate first line pharmacologic treatment (95). The UK NICE guidelines also suggest that metformin as a first-line medication for patients with GDM who require pharmacologic therapy (121).

Glyburide and Other Agents

Glyburide is the only sulfonylurea that has been studied in a large, randomized trial in individuals with GDM. It was approved by the 5th International Workshop and IADPSG as a possible alternative to insulin in individuals with GDM due to several RCTs (122). The dose ranges from 2.5-20 mg daily in divided doses.

Glyburide exposure in most RCTs is limited to the second and third trimesters, so the effect on embryogenesis was not studied, but there are no convincing reports that it is a teratogen. Due to its peak at 3-4 hours, many individuals have inadequate control of their 1- or 2-hour postprandial glucoses and then become hypoglycemic 3-4 hours later and data suggest that serum concentrations with usual doses are lower in pregnant individuals. If used, it should be given 30 mins-1-hour before breakfast and dinner and should not be given before bedtime due to the risk of nocturnal or early morning hypoglycemia considering its 3–4-hour peak (similar to regular insulin). For individuals unwilling to administer multiple daily insulin injections who have postprandial glucoses well controlled by glyburide but have fasting hyperglycemia, adding intermediate or long-acting insulin before bedtime to the glyburide can sometimes be useful. If both postprandial and fasting glucoses remain elevated, the individual should be switched to insulin.

The earliest RCTs offered glyburide as a safe alternative to insulin, without significant differences in perinatal outcomes (123). In the last 20 years, growing evidence has suggested there is increased risk of both maternal and neonatal hypoglycemia with glyburide use (109,124–127). In some trials, maternal glycemic control, macrosomia, neonatal hypoglycemia, and neonatal outcomes were not different between groups although in others, there was a significantly greater rate of macrosomic infants in the glyburide group ((123,128,129)). In a meta-analysis examining metformin versus insulin versus glibenclamide (glyburide) treatment for individuals with GDM, there were higher rates of macrosomia (risk ratio 2.62) and neonatal hypoglycemia (risk ratio 2.04) among those treated with glibenclamide compared with insulin (109). This is the same publication reviewed above that showed the increased risk of preterm birth in individuals treated with metformin compared with insulin. This meta-analysis in addition to a second meta-analysis showed significantly worse neonatal outcomes among offspring of individuals with GDM treated with glyburide compared to insulin (109,128). There were higher rates of neonatal hypoglycemia, respiratory distress syndrome, macrosomia, and birth injury without significant differences in glycemic control (124,128).

A RCT compared the efficacy of metformin with glyburide for glycemic control in GDM (124). In the individuals who achieved adequate glycemic control, the mean glucose levels were not statistically different between the two groups. However, 26 individuals in the metformin group (34.7%) and 12 individuals in the glyburide group (16.2%) did not achieve adequate glycemic control and required insulin therapy (p=0.01). Thus, in this study, the need for insulin supplementation with metformin was twice as high as the failure rate of glyburide when used in the management of GDM (124). These findings are consistent with the general finding that approximately, 15% of individuals will fail maximum dose glyburide therapy and need supplemental insulin, especially if dietary restriction is not carefully followed. Although it was initially thought not to appreciably cross the placenta or significantly affect fetal insulin levels, examination using HPLC mass spectrometry suggested a modest amount of glyburide does cross (114).

There is not sufficient information available on thiazolidinediones, meglitinides, dipeptidyl peptidase IV (DPP-4) inhibitors, glucagon like peptide 1 (GLP-1) agonists, sodium-glucose transport protein 2 (SGLT-2) inhibitors, and such agents should only be used in the setting of approved clinical trials as their teratogenic potential is unknown. Acarbose was studied in two very small studies in individuals with GDM and given its minimal GI absorption is likely to be safe, but GI side effects are often prohibitive (130). Safety studies of GLP-1 agonists and SGLT2 inhibitors have shown potential teratogenicity and adverse pregnancy outcomes mostly derived from animal studies (131). Limited observational studies in humans have not shown significant adverse outcomes with periconceptional use of GLP-1 agonists (132,133). The effect of use in later pregnancy and for gestational diabetes is unknown. Additional information about the use of these medications outside of pregnancy can be found in the chapter entitled “Oral and Injectable (Non-Insulin) Pharmacological Agents for the Treatment of Type 2 Diabetes” in the Diabetes Mellitus and Carbohydrate Metabolism section of Endotext (134).

FETAL SURVEILLANCE AND DELIVERY OPTIONS IN GESTATIONAL DIABETES

Individuals with GDM who require pharmacotherapy but do not have other comorbidities should initiate once or twice weekly antenatal fetal surveillance at 32 weeks gestation (135). There is no consensus regarding antepartum testing in individuals with diet-controlled GDM (1). For those individuals with diet-controlled GDM extending pregnancy beyond 40 weeks gestation, consideration could be made to initiate antenatal testing (135).

An ultrasound for growth to look for head to body disproportion (large abdominal circumference compared to the biparietal diameter) and evidence of LGA should be considered at ~29-32 weeks (1). The documented risks associated with attempted vaginal delivery with a fetal estimated weight >4500 gm in the setting of pregestational diabetes have resulted in a reasonable practice of offering cesarean delivery (136). This recommendation is extended to those with GDM and a fetal estimated weight >4500 gm (1). Nevertheless, a Cochrane review found insufficient evidence in using fetal biometry to assist in guiding the medical management of GDM to improve either perinatal or maternal health outcomes (137).

When GDM is well-controlled with either diet or medications, delivery <39 weeks gestation is not warranted. Delivery is usually recommended by 40 6/7 weeks for uncomplicated diet-controlled GDM and by 39 6/7 weeks for well-controlled GDM on medication(138). Earlier delivery should be considered with suboptimal glucose control or other complicating factors such as hypertension (138). The framework of evaluating the risks and benefits of induction of labor to expectant management in both high-risk and low-risk individuals has shifted from a historical lens of induction of labor compared to spontaneous labor; multiple studies have demonstrated no increased risk of cesarean delivery with induction of labor <40 weeks gestation (139–141).

POSTPARTUM ISSUES IN WOMEN WITH GESTATIONAL DIABETES

Re-Evaluating Glucose Tolerance Postpartum and Future Risk of Diabetes

Identification of poor glycemic control in pregnancy to predict risk for T2DM was present in the earliest screening and diagnostic strategies. Up to 70% of individuals with GDM are estimated to ultimately develop T2DM within 20-30 years after delivery. Differentiation of GDM from previously undiagnosed T2DM should be performed via a 75-gram 2-hour GTT within 4-12 weeks postpartum as recommended by the ADA, Canadian Diabetes Association (CDA), Fifth International Workshop, and ACOG since most individuals with impaired glucose intolerance will be missed if only a FBG is checked (142). A 2-hour value of at least 200 mg/dl establishes a diagnosis of diabetes and a 2-hour value of at least 140 mg/dl but less than 200 mg/dl makes the diagnosis of impaired glucose tolerance. Additionally, individuals who have been diagnosed with GDM should be screened at least every 3 years for overt diabetes (37).

An alternative approach that is endorsed by ACOG in a 2024 Clinical Practice Update is to perform the 75-gram GTT on postpartum day 2 prior to hospital discharge, since completion of postpartum screening is historically low (54). A large series of ~23,000 individuals who received lab testing through Quest diagnostics suggested that only 19% of individuals receive postpartum diabetes testing within a 6-month period (143). Waters et al. demonstrated a negative GTT prior to hospital discharge excluded T2DM diagnosis at 4-12 weeks postpartum (144). Werner et al. showed similar diagnostic value of 2-day postpartum GTT to the standard 4-12 weeks postpartum GTT (145)

A weight loss program consisting of diet and exercise should be instituted for individuals with GDM to improve their insulin sensitivity and hopefully to prevent the development of T2DM (146). Hyperglycemia generally resolves in most individuals during this interval but up to 10% of patients will fulfill criteria for T2DM. At the minimum, a fasting blood glucose should be done to determine if the woman has persistent diabetes (glucose >125 mg/dl) or impaired fasting glucose (glucose ≥ 100 mg/dl). Of note, breastfeeding has been shown to improve insulin resistance and glucose values in postpartum individuals with GDM (147,148).

Utility of using the A1c postpartum to predict the subsequent occurrence of T2DM in individuals with a history of GDM has not been studied extensively and may be affected by glycemic control during pregnancy if done before 3 months postpartum (149). A study looking at utility of using A1c vs 2h GTT vs FPG for screening of individuals with recent GDM showed that A1c and A1c plus FPG did not have the sensitivity and specificity to diagnose impaired carbohydrate metabolism postpartum (150,151). The importance of diagnosing impaired glucose intolerance lies in its value in predicting the future development of T2DM. In one series which mainly studied Hispanic individuals, a diagnosis of impaired glucose tolerance was the most potent predictor of the development of T2DM in individuals with a history of GDM; 80% of such women developed diabetes in the subsequent 5-7 years (152). Intensified efforts promoting diet, exercise and weight loss should be instituted in these individuals.

Other studies have shown other risk factors for development of prediabetes and/or T2DM after GDM including earlier diagnosis of GDM in pregnancy, insulin therapy during pregnancy, and BMI (64,153,154). A study in Italy showed pre-pregnancy BMI and PCOS as strong predictors of postpartum impaired glucose tolerance (155,156). A1c within 12 months postpartum may be useful in addition to GTT to diagnose some women with history of GDM and normal glucose tolerance. A study of 141 individuals in Spain with recent GDM found that 10% had normal glucose tolerance, normal FPG, and isolated A1c 5.7-6.4% suggesting that A1c is a useful tool to diagnose prediabetes in individuals with a history of GDM with normal glucose tolerance postpartum (156). Interestingly, in this study the group of individuals with isolated A1c 5.7-6.4% with normal glucose tolerance and normal FPG were more likely to be non-Hispanic White and more likely had higher LDL-C values. A1c is a sensitive test in detecting prediabetes and overt diabetes in postpartum individuals with history of GDM (157).

The TRIPOD study demonstrated that the use of a thiazolidinedione postpartum in individuals with a history of GDM and persistent impaired glucose intolerance decreased the development of T2DM (158). The rate of T2DM in the 133 individuals randomized to troglitazone was 5.4% versus 12.1% in the 133 individuals randomized to placebo at a median follow-up of 30 months (159). The protection from diabetes was closely related to the degree of reduction of insulin secretion three months after randomization and persisted 8 months after the medication was stopped. In the PIPOD study, use of Pioglitazone to the same high-risk patient group stabilized previously falling β-cell function and revealed a close association between reduced insulin requirements and low risk of diabetes (7,67). However, using thiazolidinediones for the purpose of preventing the development of T2DM in individuals with a history of GDM has not been recommended. The Diabetes Prevention Program (DPP) Trial analyzed their data in individuals with a history of GDM (66). A total of 349 individuals had a history of GDM, and such a history conferred a 74% increased risk for the development of T2DM compared with individuals without a history of GDM. In the placebo arm, individuals developed T2DM at an alarming rate of 17% per year but this rate was cut in half by either use of metformin or diet and exercise.

The DPP, TRIPOD, and PIPOD studies support clinical management that focuses on identifying individuals who meet criteria for metabolic syndrome, achieving postpartum weight loss, and instituting aggressive interventions beginning with lifestyle changes to decrease insulin resistance for the primary prevention of T2DM. Individuals with a history of GDM who display normal testing postpartum should undergo lifestyle interventions for postpartum weight reduction and receive repeat testing at least every 3 years (37). For individuals who may have subsequent pregnancies, screening more frequently has the advantage of detecting abnormal glucose metabolism before the next pregnancy to ensure preconception glucose control (1).

Breastfeeding

Breastfeeding should be encouraged in all individuals with a history of GDM for improving maternal and offspring health outcomes. Lactation completes the reproductive cycle and is associated with significant short- and long-term cardiometabolic benefits for both mother and infant, with most demonstrating a dose-dependent relationship (160). Professional society recommendations recommend exclusive breastfeeding for the first 6 months of life, then ongoing breastfeeding with complementary foods through the second year or as long as desired by both mother and child (160–162).

Initially the correlation between breastfeeding and reduced incidence of T2DM were based on self-reported lactation status and diabetes diagnoses. Subsequent larger studies have confirmed this relationship. Over 1200 individuals who had at least 1 live birth and reported lactation duration were followed in a community-based prospective study over 30 years, with diabetes screening performed up to 7 times (CARDIA study) (163). Not only was there a three-fold increased incidence of T2DM in those with no breastfeeding compared to any breastfeeding, but the relative hazard was also graded based on duration of breastfeeding (163). These findings were also reported in the most recent meta-analysis evaluating the relationship between lactation and maternal risk of T2DM (164).

For children, breastmilk intake also appears to decrease the risk of developing obesity and impaired glucose tolerance (165). In the large EPOCH study (Exploring Perinatal Outcomes Among Children Study), offspring of individuals with diabetes (primarily GDM) who were breastfed for at least 6 months had a slower BMI growth trajectory during childhood and a lower childhood BMI than those who were not breastfed (166). There is a growing literature suggesting that some of the protective benefits on childhood obesity and programming the infant immune system from breast milk may be influenced by appetite regulatory hormones, biomarkers of oxidative stress and inflammation, and the milk microbiome (167–170).

Contraception

Discussing contraception and family planning during pregnancy is an effective way to promote safe pregnancy interval, with optimal outcomes observed when delivery and conception are at least 18 months apart (171). For individuals with GDM, the postpartum and inter-pregnancy periods offer a tremendous opportunity to employ diet, lifestyle, and other therapeutic changes to reduce the risk of subsequent GDM or T2DM (171). All pharmacologic options for contraception are considered safe in the setting of recent or remote GDM, though estrogen-containing methods should be delayed until ≥21 days postpartum to reduce the risk of thromboembolism (172). Estrogen may negatively impact breast milk production, so consideration of infant feeding method should also be weighed against initiation. We recommend a patient-centered approach to counseling and selecting a contraceptive method.

There is limited data on the influence of various contraceptive methods on long-term risk of T2DM, insulin sensitivity, glycemic control, weight gain, and hypercholesterolemia (173). Extensive research evaluating these relationships concludes the adverse outcomes observed with methods such as Depo-Provera in the GDM population are more closely associated with initial BMI and pregnancy weight gain than with GDM.

CONCLUSION

Different organizations recommend different screening and diagnostic strategies for GDM reflecting variations in geographic settings. Treatment with lifestyle or medication to achieve glycemic targets improves obstetric and perinatal outcomes. Due to the obesity epidemic, the incidence of GDM is only expected to rise, with subsequent or eventual T2DM diagnosis increasing accordingly. Further investigation on the benefits of specific pharmacologic therapies and glucose monitoring strategies with CGM are ongoing.

The development of T2DM in individuals with a history of GDM as well as obesity and glucose intolerance in the offspring of those with preexisting DM or GDM set the stage for a perpetuating metabolic cycle that must be aggressively addressed with effective primary prevention strategies that begin in-utero. Pregnancy is a unique opportunity to implement strategies to improve the mother’s lifetime risk for adverse cardiometabolic health outcomes in addition to that of her offspring and offers the potential to decrease the intergenerational risk of obesity, diabetes, and other adverse cardiometabolic outcomes.