Gestational Diabetes

Gestational diabetes mellitus (GDM) is a form of diabetes diagnosed during pregnancy, usually in the second or third trimester, that was not clearly present before conception. It is distinguished from pre-existing diabetes and from previously unrecognized diabetes first detected during pregnancy at levels meeting standard diabetes diagnostic criteria. Most affected people have no noticeable symptoms, so identification generally depends on blood glucose testing rather than symptoms alone. (diabetesjournals.org)

The condition develops when pancreatic insulin secretion cannot adequately compensate for the increased insulin resistance associated with pregnancy. Elevated maternal glucose is associated with excessive fetal growth and other pregnancy complications. Although glucose regulation commonly returns to normal after delivery, a history of gestational diabetes is associated with a substantially increased risk of subsequent type 2 diabetes. (niddk.nih.gov)

Gestational diabetes connects maternal metabolism, fetal development, obstetric care, and long-term disease prevention. Its reported frequency depends partly on the population studied and the diagnostic criteria used. Understanding the condition therefore requires attention to its biological mechanisms, screening and diagnosis, maternal and infant outcomes, clinical management, postpartum course, and the research underlying diagnostic and treatment practices. (niddk.nih.gov)

Definition and Clinical Classification

Gestational diabetes mellitus (GDM) is a clinical category of diabetes mellitus defined partly by its relationship to pregnancy. The American Diabetes Association’s (ADA) Standards of Care in Diabetes—2026 defines it as diabetes diagnosed in the second or third trimester that was not clearly overt diabetes before gestation, excluding other forms of diabetes occurring during pregnancy, such as type 1 diabetes. This definition distinguishes pregnancy-associated glucose intolerance from established diabetes complicating pregnancy. It does not establish that the underlying metabolic abnormality began on the date of diagnosis: previously unrecognized hyperglycemia or impaired capacity to compensate for insulin resistance may precede pregnancy. (diabetesjournals.org)

The historical definition was broader. In 1979, the National Diabetes Data Group described gestational diabetes as carbohydrate intolerance of variable severity beginning or first recognized during pregnancy. Under this approach, the timing of recognition was sufficient for classification, regardless of whether glucose intolerance had actually existed beforehand. It therefore grouped relatively mild pregnancy-associated hyperglycemia with previously undiagnosed diabetes severe enough to meet ordinary diabetes criteria outside pregnancy. The contemporary distinction addresses this limitation by separating GDM from overt diabetes first detected during gestation; neither the historical nor the contemporary designation, by itself, guarantees that glucose regulation will return to normal after delivery. (niddk.nih.gov)

Pre-existing, or pregestational, diabetes denotes diabetes present before conception, including type 1 diabetes and type 2 diabetes. A person with established type 1 diabetes remains classified as having type 1 diabetes during pregnancy, rather than acquiring a GDM diagnosis because insulin requirements change. Likewise, known type 2 diabetes remains type 2 diabetes complicating pregnancy. Previously unrecognized diabetes discovered during prenatal assessment belongs to a separate category when it meets standard diabetes diagnostic criteria. The date of detection does not necessarily identify the date of disease onset, and overt hyperglycemia alone does not establish whether the underlying diabetes is type 1, type 2, or another specific form. (diabetesjournals.org)

The World Health Organization (WHO) formalized this separation in its 2013 guideline, classifying hyperglycemia first detected during pregnancy as either diabetes mellitus in pregnancy or gestational diabetes mellitus. The former, often termed overt diabetes in pregnancy, meets diabetes thresholds used outside pregnancy: examples include fasting plasma glucose of at least 7.0 mmol/L (126 mg/dL), or a two-hour value of at least 11.1 mmol/L (200 mg/dL) after a 75-g oral glucose load. GDM occupies a lower, pregnancy-specific diagnostic range. Unlike the ADA’s second- or third-trimester wording, the WHO’s 2013 definition permits GDM classification at any time during pregnancy. These formulations therefore differ in timing, while both distinguish GDM from diabetes-level hyperglycemia first recognized during gestation. (ncbi.nlm.nih.gov)

Clinical records also use A1GDM and A2GDM to describe treatment requirements. A1GDM denotes GDM managed without glucose-lowering medication, generally through nutritional measures and physical activity; A2GDM denotes GDM requiring medication, including insulin or an oral glucose-lowering agent. These are management categories, not separate biological diseases or synonyms for type 1 and type 2 diabetes. For example, a pregnancy initially recorded as A1GDM may subsequently be recorded as A2GDM when medication is introduced, without changing the underlying diagnosis of GDM. Insulin treatment alone does not reclassify GDM as type 1 diabetes. (pmc.ncbi.nlm.nih.gov)

Epidemiology and Risk Factors

Reported prevalence of gestational diabetes depends on both the population studied and how cases are identified. Lower glucose thresholds, or diagnostic rules requiring only one abnormal value rather than several, identify more affected pregnancies. Universal testing and testing restricted to selected risk groups also produce different case counts. Maternal age, prepregnancy body mass, and the population’s underlying frequency of glucose intolerance influence the estimates independently of testing methods. Consequently, an increase after a change in diagnostic criteria does not necessarily represent an equivalent increase in underlying metabolic disease. Boyd E. Metzger and Thomas A. Buchanan’s chapter in Diabetes in America, third edition, specifically notes that adoption of the International Association of Diabetes and Pregnancy Study Groups criteria increases measured prevalence. (pubmed.ncbi.nlm.nih.gov)

The denominator and data source are equally important. Estimates from screened pregnancy cohorts, hospital delivery records, and birth certificates are not interchangeable: they differ in coverage, clinical detail, and completeness of recording. Statistics restricted to live births describe pregnancies resulting in a registered live birth, rather than all pregnancies, and do not include pregnancies ending in miscarriage or stillbirth. Birth-certificate recording may also omit a diagnosis made during prenatal care. Such figures are therefore measures of recorded disease within a defined surveillance system, not direct measurements of every affected pregnancy. (niddk.nih.gov)

A dated national example comes from the United States National Vital Statistics System. An analysis published by Joyce A. Martin and Elizabeth C. W. Gregory on January 6, 2023, reported that the percentage of mothers giving birth with gestational diabetes recorded on the birth certificate increased from 6.0% in 2016 to 8.3% in 2021—an increase of 2.3 percentage points. Increases occurred in every maternal age group. In 2021, the recorded percentage was 2.7% among mothers younger than 20 years and 15.6% among those aged 40 years or older, nearly a sixfold difference. Nationwide birth-certificate information on gestational diabetes first became available in 2016; the report explicitly cautioned that the condition might be underreported. (cdc.gov)

Established risk factors include higher prepregnancy body mass, obesity, older maternal age, a family history of diabetes—particularly type 2 diabetes in a parent or sibling—and gestational diabetes in a previous pregnancy. Pre-existing glucose intolerance and polycystic ovary syndrome are additional recognized risk factors. Higher adiposity can contribute through insulin resistance already present before pregnancy, increasing the insulin secretion required to maintain normal glucose concentrations. Family history is consistent with inherited susceptibility, although it does not isolate genetic effects from shared environmental influences. These factors increase probability rather than determine diagnosis: gestational diabetes also occurs in women without recognized risk factors. (niddk.nih.gov)

Population comparisons require similar caution. For example, age-adjusted hospital-discharge estimates from 23 U.S. states in 2008 recorded 4.40 cases per 100 deliveries among non-Hispanic white women, 5.30 among non-Hispanic black women, 7.02 among Hispanic women, and 8.14 among Asian women. These broad categories describe observed population differences, not a single ancestry-specific biological cause. Their interpretation must account for maternal age, adiposity, underlying diabetes susceptibility, and differences in ascertainment. Because diagnosis depends on access to and completion of testing, incomplete screening can lower recorded prevalence without demonstrating lower underlying risk. (niddk.nih.gov)

Pregnancy-Related Metabolic Mechanisms

During the second and third trimesters of pregnancy, maternal tissues become progressively less responsive to insulin. This physiological insulin resistance increases the insulin secretion required to maintain normal blood glucose. Pancreatic beta cells ordinarily compensate by increasing their response to glucose, so insulin resistance alone does not constitute gestational diabetes. A 1991 longitudinal study by Patrick M. Catalano and colleagues illustrates the scale of this adaptation: in six nonobese women examined before conception, at 12–14 weeks, and at 34–36 weeks, insulin sensitivity decreased by 56%, while first- and second-phase insulin release during intravenous glucose testing increased approximately three- to three-and-a-half-fold. These measurements describe a small physiological study rather than fixed changes occurring in every pregnancy. (niddk.nih.gov)

The placenta contributes to this changing metabolic environment through its endocrine activity. Placental growth hormone and human placental lactogen participate in a network of pregnancy-related hormonal influences, rather than acting as a single demonstrated cause of gestational diabetes. Experimental evidence supports an insulin-antagonistic action of placental growth hormone: in a 2002 study by Linda A. Barbour and colleagues, transgenic mice expressing this human hormone at concentrations comparable to late human pregnancy developed marked insulin resistance, with fasting insulin approximately four times that of controls. This establishes an effect in an animal model, not its quantitative contribution in pregnant humans. Pregnancy-related hormonal changes also coexist with maternal weight gain and other metabolic changes that influence insulin responsiveness. (pubmed.ncbi.nlm.nih.gov)

Two important consequences concern peripheral glucose disposal and endogenous glucose production. Reduced insulin sensitivity means that peripheral tissues, particularly skeletal muscle, take up less glucose for a given insulin concentration. Meanwhile, basal glucose production increases as pregnancy advances. Normally, insulin restrains hepatic glucose release; inadequate suppression contributes to elevated fasting glucose, whereas reduced peripheral disposal and an insufficient insulin response contribute to post-meal hyperglycemia. Catalano and colleagues’ 1993 study followed six control participants and ten women who developed gestational diabetes from before conception into late pregnancy. Basal endogenous glucose production increased during gestation, but the gestational-diabetes group showed less suppression of production during insulin infusion in late pregnancy. The findings identify a hepatic component alongside peripheral insulin resistance and impaired secretion. (pubmed.ncbi.nlm.nih.gov)

Gestational diabetes develops when beta-cell compensation is insufficient relative to the prevailing insulin resistance; absolute insulin concentrations may therefore be elevated yet still inadequate. Such impaired compensation can precede pregnancy rather than arise entirely from placental influences. The relative contributions vary between individuals. In a 2016 study led by Camille E. Powe, glucose-test-derived indices were assessed in 809 women at 24–30 weeks, of whom 67 developed gestational diabetes. Using values below the 25th percentile of the normal-glucose-tolerance group to define defects, approximately 51% of affected participants had predominantly impaired insulin sensitivity, 30% predominantly impaired secretion, and 18% both. These were research-defined physiological subgroups, not universally applicable clinical categories. (niddk.nih.gov)

Maternal glucose crosses the placenta and stimulates insulin secretion by the fetal pancreas; maternal insulin does not ordinarily cross into the fetal circulation. Consequently, fetal hyperinsulinemia reflects the fetus’s own response to increased nutrient exposure. Fetal insulin promotes growth and fat deposition, providing a biological connection between maternal hyperglycemia and increased neonatal adiposity. A 2009 analysis from the Hyperglycemia and Adverse Pregnancy Outcome study found graded associations of maternal glucose and cord-blood C-peptide—a marker of fetal insulin secretion—with neonatal skinfold thickness and estimated body-fat percentage. These associations persisted after adjustment for factors including maternal body mass index and gestational age, supporting the proposed glucose–fetal insulin–growth pathway without making excessive growth inevitable in an individual pregnancy. (diabetesjournals.org)

Clinical Presentation and Screening

Gestational diabetes mellitus (GDM) usually produces no recognizable symptoms and is commonly detected through scheduled blood testing rather than a symptom-driven investigation. When symptoms occur, they may include increased thirst or more frequent urination, but these have limited specificity during pregnancy: similar complaints can occur without diabetes. Their presence does not establish GDM, and their absence does not exclude it. This largely asymptomatic presentation is the principal reason screening recommendations include pregnant people who feel well. (niddk.nih.gov)

The usual screening window is 24–28 weeks of gestation. Pregnancy-associated insulin resistance becomes more pronounced as pregnancy advances; GDM emerges when pancreatic insulin secretion cannot adequately compensate. Consequently, normal glucose measurements early in pregnancy do not exclude later gestational hyperglycemia. The American Diabetes Association’s (ADA) Standards of Care in Diabetes—2026 recommends screening at 24–28 weeks in pregnant individuals not already identified as having diabetes or high-risk abnormal glucose metabolism earlier in the pregnancy. The U.S. Preventive Services Task Force (USPSTF), in its recommendation dated August 10, 2021, recommends screening asymptomatic pregnant people at 24 weeks or later, assigning a grade B recommendation. Its finding of insufficient evidence for GDM screening before 24 weeks concerns the balance of benefits and harms of early screening, not whether previously unrecognized diabetes can be detected early. (niddk.nih.gov)

Early testing and routine later GDM screening have different purposes. Under the ADA’s 2026 guidance, testing before 15 weeks, at the first prenatal visit if preconception testing has not occurred, is recommended for individuals with diabetes risk factors and may be considered for everyone. This testing uses standard diagnostic criteria to identify previously unrecognized diabetes mellitus, rather than automatically classifying every early abnormal result as GDM. The ADA separately identifies early abnormal glucose metabolism below diabetes thresholds as a marker of increased pregnancy risk and later GDM risk. It also cautions that the usual GDM testing criteria were not derived from data in the first half of pregnancy and should not be used for early screening. A negative early diabetes screen is followed by routine GDM screening at 24–28 weeks. (diabetesjournals.org)

Universal screening applies testing to the eligible pregnant population irrespective of recorded risk factors; risk-factor-based screening selects individuals using their history and characteristics. The USPSTF’s 2021 recommendation exemplifies universal screening at 24 weeks or later. In contrast, the United Kingdom’s National Institute for Health and Care Excellence (NICE), in recommendations issued in 2015 within guideline NG3, uses risk-factor-based testing. Examples of qualifying factors include a body mass index of at least $30\ \mathrm{kg/m^2}$, a previous baby weighing at least 4.5 kg, previous GDM, or diabetes in a first-degree relative. NICE specifies a 75-g, two-hour oral glucose tolerance test at 24–28 weeks for those with qualifying factors other than previous GDM. Previous GDM prompts early glucose self-monitoring or an early tolerance test, with another tolerance test at 24–28 weeks if the early result is normal. (uspreventiveservicestaskforce.org)

Population selection is distinct from the choice of testing sequence. In a two-step pathway, an initial glucose challenge test—commonly using 50 g of glucose—measures blood glucose after one hour and does not require fasting. A positive screen ordinarily leads to a separate fasting oral glucose tolerance test (OGTT), commonly using 100 g over three hours; the initial screening result is not itself the usual diagnostic test. In a one-step pathway, the fasting 75-g OGTT is administered directly, without a preceding glucose challenge. Thus, “universal screening” does not necessarily mean a universal preliminary challenge test: it may instead mean offering the diagnostic OGTT to all eligible individuals. Numerical diagnostic thresholds are specific to the selected protocol. (niddk.nih.gov)

Diagnostic Tests and Thresholds

The one-step method uses a fasting 75-g oral glucose tolerance test (OGTT), with laboratory plasma glucose measurements before the glucose drink and at one and two hours afterward. The American Diabetes Association’s (ADA) Standards of Care in Diabetes—2026 specifies morning testing after an overnight fast of at least eight hours, ordinarily at 24–28 weeks of pregnancy. The International Association of Diabetes and Pregnancy Study Groups (IADPSG) thresholds, incorporated into the World Health Organization (WHO) classification in 2013, are fasting 5.1 mmol/L (92 mg/dL), one hour 10.0 mmol/L (180 mg/dL), or two hours 8.5 mmol/L (153 mg/dL). Meeting or exceeding any one threshold is sufficient; all three measurements need not be abnormal. (diabetesjournals.org)

WHO distinguishes this diagnosis from diabetes mellitus in pregnancy, sometimes termed overt diabetes first detected during pregnancy. Its gestational-diabetes ranges are fasting glucose 5.1–6.9 mmol/L (92–125 mg/dL) or two-hour glucose 8.5–11.0 mmol/L (153–199 mg/dL), alongside the one-hour threshold above. Fasting glucose of at least 7.0 mmol/L (126 mg/dL), or two-hour glucose of at least 11.1 mmol/L (200 mg/dL) after a 75-g load, instead meets WHO criteria for diabetes mellitus in pregnancy. Random plasma glucose of at least 11.1 mmol/L (200 mg/dL) also qualifies when accompanied by diabetes symptoms. There is no corresponding established one-hour OGTT threshold for overt diabetes. These distinctions prevent more marked hyperglycemia from being classified solely as gestational diabetes. (ncbi.nlm.nih.gov)

The two-step method separates screening from diagnostic testing. First, a nonfasting 50-g glucose challenge measures plasma glucose after one hour. A result at or above the selected screening cutoff—130, 135, or 140 mg/dL (7.2, 7.5, or 7.8 mmol/L)—triggers a fasting 100-g, three-hour OGTT rather than automatically establishing the diagnosis. The Carpenter–Coustan thresholds are fasting 95 mg/dL (5.3 mmol/L), one hour 180 mg/dL (10.0 mmol/L), two hours 155 mg/dL (8.6 mmol/L), and three hours 140 mg/dL (7.8 mmol/L). The usual rule requires at least two values meeting or exceeding their thresholds. However, the ADA’s 2026 table notes that the American College of Obstetricians and Gynecologists permits diagnosis using one elevated value, so the number-of-values rule must be specified when reporting results. (diabetesjournals.org)

Other measurements do not provide an interchangeable substitute for the OGTT. A fasting measurement can itself satisfy the one-step fasting criterion, but a value below that cutoff does not exclude an abnormal post-load result. Random glucose lacks the standardized fasting state, glucose dose, and sampling interval. Urine glucose is not a plasma measurement or a diagnostic endpoint in either protocol; NICE’s Diabetes in pregnancy guideline treats specified degrees of glycosuria as a reason for further testing, not as a diagnosis. Glycated hemoglobin (HbA1c) reflects longer-term glycemia rather than the timed glucose response, and pregnancy-associated changes in red-cell turnover alter its relationship with glucose concentrations. NICE does not use HbA1c as a substitute for its diagnostic OGTT. (uspreventiveservicestaskforce.org)

Diagnostic cutoffs also differ from treatment targets. The former classify laboratory results obtained under a defined glucose-loading protocol; the latter assess fasting and post-meal glucose during ongoing management. For example, the ADA’s 2026 management standards specify fasting glucose below 95 mg/dL (5.3 mmol/L), one-hour post-meal glucose below 140 mg/dL (7.8 mmol/L), or two-hour post-meal glucose below 120 mg/dL (6.7 mmol/L). An ordinary meal is not equivalent to a standardized 75-g or 100-g glucose load, so these numbers cannot be exchanged with OGTT thresholds. (pmc.ncbi.nlm.nih.gov)

Differences in glucose dose, screening cutoffs, diagnostic thresholds, and the required number of abnormal values change how many pregnancies receive the diagnosis. In Teresa A. Hillier and colleagues’ 2021 randomized trial, involving 23,792 women in two health systems, gestational diabetes was diagnosed in 16.5% assigned to one-step screening versus 8.5% assigned to two-step screening. The trial found no significant between-group differences in its other primary maternal and perinatal outcomes. These percentages therefore illustrate the effect of the testing strategy on diagnostic frequency, rather than a difference in underlying disease occurrence between independently sampled populations. (nejm.org)

Maternal, Fetal, and Neonatal Outcomes

Gestational diabetes mellitus (GDM) is associated with higher frequencies of maternal hypertensive disorders, including gestational hypertension and preeclampsia, and with cesarean delivery. Preeclampsia involves hypertension arising after 20 weeks of pregnancy with proteinuria or other evidence of organ dysfunction; it is not simply an elevated glucose concentration. These associations describe increased probabilities, not inevitable complications. Maternal obesity and other characteristics can contribute to risk alongside hyperglycemia. The 2008 Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study found graded associations between maternal glucose and several adverse outcomes after adjustment for potential confounders, rather than a sharp glucose level separating uncomplicated from complicated pregnancies. Consequently, a GDM diagnosis alone does not specify an individual pregnancy’s outcome. (nichd.nih.gov)

Excessive fetal growth is described using two distinct measurements. Large for gestational age (LGA) generally means birth weight at or above the 90th percentile for the infant’s gestational age on the reference chart used. Macrosomia instead denotes an absolute birth weight, commonly 4,000 g or 4,500 g, irrespective of gestational age; definitions differ between studies and clinical settings. Thus, an infant can be LGA without meeting a macrosomia threshold. Excessive size increases the likelihood of difficult delivery, particularly shoulder dystocia, in which the shoulders do not deliver normally after the head has emerged and additional obstetric maneuvers are required. Associated injuries include clavicular or humeral fractures and brachial plexus injury, which can impair arm movement. Neither macrosomia nor GDM makes shoulder dystocia certain, and shoulder dystocia can occur in infants without excessive birth weight. (studylib.net)

The variability and modifiability of these outcomes are illustrated by Mark B. Landon and colleagues’ 2009 trial involving 958 women with mild GDM. LGA birth occurred in 7.1% of the treatment group versus 14.5% of the usual-care group, while shoulder dystocia occurred in 1.5% versus 4.0%. Cesarean delivery rates were 26.9% versus 33.8%, and the combined frequency of preeclampsia and gestational hypertension was 8.6% versus 13.6%. These figures are specific to the trial population, not universal estimates for all GDM pregnancies. Although several secondary outcomes improved, the primary composite of perinatal death and selected neonatal complications did not differ significantly between groups. (nejm.org)

Neonatal hypoglycemia reflects the transition from placental nutrition to independent glucose regulation. Maternal glucose crosses the placenta, and elevated fetal glucose exposure stimulates fetal insulin production. At birth, separation from the placenta abruptly ends the maternal glucose supply, while increased insulin secretion can persist, lowering the newborn’s blood glucose. GDM studies also report respiratory distress, preterm birth, hyperbilirubinemia or jaundice, and neonatal intensive-care admission. Respiratory morbidity is influenced by gestational age and delivery circumstances as well as the diabetic pregnancy; it should not automatically be attributed to glucose exposure alone. Outcome definitions also matter: studies may count clinically diagnosed hypoglycemia, specified laboratory glucose values, or treatment requirements, which are not interchangeable endpoints. (ncbi.nlm.nih.gov)

Typical GDM developing later in pregnancy must be distinguished from type 1 diabetes, type 2 diabetes, or previously unrecognized overt diabetes already present during early embryonic development. Major organs begin forming during the first eight weeks of pregnancy, when substantial hyperglycemia can increase the risk of congenital malformations involving structures such as the heart, brain, and kidneys. Hyperglycemia arising only later does not represent the same exposure during organ formation: its characteristic outcome profile centers on fetal growth, delivery complications, and neonatal metabolic adaptation. The date diabetes is first detected, however, does not by itself establish when hyperglycemia began, so malformation risks associated with pre-existing diabetes cannot simply be assigned to every pregnancy diagnosed with GDM. (niddk.nih.gov)

Glucose Monitoring and Nonpharmacological Management

Care commonly involves obstetric clinicians, diabetes specialists, nurses or diabetes educators, and registered dietitian nutritionists. Their complementary tasks include coordinating pregnancy care, teaching glucose-meter use, reviewing glucose records, and developing an individualized eating plan. Education links day-to-day measurements with food choices and activity rather than treating glucose testing as an isolated procedure. Multidisciplinary care can be organized within a specialist service or through professionals working across several services. (niddk.nih.gov)

Medical nutrition therapy aims to moderate glucose excursions while supplying sufficient energy and nutrients for maternal health and fetal development. Food type, quantity, and timing influence post-meal glucose concentrations; assessment therefore considers eating patterns rather than simply eliminating carbohydrate. The American Diabetes Association’s (ADA) 2026 standards describe individualized planning with a dietitian and assessment of gestational weight gain in relation to prepregnancy weight. Physical activity is another component of glucose management, although the evidence does not establish one universally optimal exercise program for GDM. (niddk.nih.gov) Dietary flexibility was illustrated by Teri L. Hernandez and colleagues’ 2023 CHOICE randomized trial: provided diets containing 60% versus 40% of energy from carbohydrate produced similar maternal 24-hour glycemia and newborn adiposity. This comparison concerned structured diets with differing carbohydrate and fat content, not unrestricted eating. (diabetesjournals.org)

Capillary monitoring uses a finger-prick blood sample and a glucose meter, with fasting and post-meal readings providing complementary information about overnight and meal-related glycemia. Recorded results allow the clinical team to assess whether the management plan is achieving its intended effect. (niddk.nih.gov) The ADA’s 2026 management targets are fasting glucose below 95 mg/dL (5.3 mmol/L), and either one-hour postprandial glucose below 140 mg/dL (7.8 mmol/L) or two-hour glucose below 120 mg/dL (6.7 mmol/L). These are treatment targets, not the laboratory oral glucose tolerance test thresholds used for diagnosis. (diabetesjournals.org)

Continuous glucose monitoring (CGM) measures glucose in interstitial fluid, generating a profile that includes time within, above, or below a specified range. Interstitial readings can lag behind blood glucose during rapid changes. The ADA’s 2026 standards recommend CGM in pregnancy with type 1 diabetes, but identify insufficient evidence for universal use in GDM. GDM-specific trials have reported differing results: one found improved time in range, whereas another found no significant improvement in its primary composite perinatal outcome. Optimal GDM-specific time-in-range targets also remain undefined; evidence from type 1 diabetes pregnancy cannot simply be transferred to GDM. (diabetesjournals.org)

Telehealth supports monitoring through electronic glucose records, remote consultations, and clinician feedback between visits. Its evidence concerns a care-delivery package rather than a glucose-measuring device alone. Weihua Xie and colleagues’ 2020 meta-analysis included 32 randomized trials and 5,108 participants with GDM. Compared with standard care, telemedicine-supported care improved pooled glucose measures and reduced cesarean delivery risk, with a relative risk of 0.82. Interpretation depends on the intervention: studies used web systems, apps, connected devices, and messaging services; 21 of the 32 trials were conducted in China, and several glucose outcomes showed substantial heterogeneity. (pmc.ncbi.nlm.nih.gov)

Medication and Obstetric Care

Medication is introduced when medical nutrition therapy and physical activity do not sufficiently control fasting or post-meal glucose concentrations. Its purpose is to reduce maternal hyperglycemia and consequent fetal glucose exposure, rather than to replace nonpharmacological management. The American Diabetes Association’s (ADA) Standards of Care in Diabetes—2026 identifies insulin as the preferred medication for GDM. Basal insulin addresses glucose elevations between meals and overnight, whereas mealtime insulin addresses postprandial elevations; the regimen depends on the observed glucose pattern. Human insulin preparations have not been demonstrated to cross the placenta. Treatment requires continued glucose monitoring because insulin can cause maternal hypoglycemia, and requirements change during pregnancy. (diabetesjournals.org)

Metformin and glyburide, also called glibenclamide, are oral alternatives with different mechanisms: metformin principally reduces hepatic glucose production, whereas glyburide stimulates pancreatic insulin secretion. Both cross the placenta. The ADA’s 2026 standards do not recommend either as first-line treatment, citing fetal exposure, incomplete achievement of glycemic goals in some patients, and uncertainty about long-term offspring safety. Glyburide concentrations in umbilical cord plasma are approximately 50–70% of maternal concentrations; comparative evidence associates its use with more neonatal hypoglycemia and macrosomia than insulin or metformin. These findings distinguish glyburide from metformin rather than establishing equivalent safety for all oral agents. (s3.amazonaws.com)

Guidelines differ in their positioning of metformin. The Society for Maternal-Fetal Medicine’s 2018 statement, reaffirmed in 2024, describes it as a reasonable first-line pharmacological alternative to insulin after nutrition therapy proves insufficient, while noting that approximately half of treated women still require insulin. The United Kingdom’s NICE guideline NG3, in recommendations issued in 2015, offers metformin when diet and exercise have not achieved glucose targets within one to two weeks. It specifies insulin when metformin is contraindicated or unacceptable, and additional insulin when metformin is insufficient. NICE also provides for immediate insulin treatment in more marked hyperglycemia or selected cases with fetal complications. These are attributed clinical pathways, not a universal medication hierarchy. (publications.smfm.org)

Obstetric monitoring distinguishes assessment of fetal growth from surveillance of fetal well-being. Ultrasonography assesses growth and amniotic fluid volume; NICE specifies growth scans every four weeks from 28 through 36 weeks. Antenatal surveillance uses tests such as fetal heart-rate recording and the biophysical profile to identify potentially compromised fetuses. Its timing and frequency depend on glucose control, medication use, additional maternal disease, and fetal findings. The American College of Obstetricians and Gynecologists’ 2021 guidance states that once- or twice-weekly surveillance may be considered from 32 weeks in medication-controlled GDM, and twice-weekly surveillance in poorly controlled GDM. It identifies no consensus on surveillance before 40 weeks for diet-controlled GDM without other complications. (nice.org.uk)

Delivery planning considers glycemic control, medication requirements, fetal growth, surveillance results, and complications such as hypertension. These factors inform the balance between continued pregnancy and planned birth; GDM alone does not establish a requirement for cesarean delivery. NICE calls for discussion of vaginal birth, induction, and cesarean delivery when ultrasound suggests macrosomia. During labor, maternal glucose is monitored because glycemic management remains relevant to neonatal hypoglycemia risk. NICE’s 2015-amended intrapartum recommendations specify hourly capillary glucose measurement and protocol-based intravenous insulin and dextrose when its labor glucose range cannot otherwise be maintained. (acog.org)

Postpartum Course and Long-Term Health

Delivery of the placenta removes a major source of hormones promoting insulin resistance, so maternal insulin requirements fall sharply after childbirth. Pregnancy-associated hyperglycemia often resolves, but normal glucose immediately after delivery does not establish that glucose regulation will remain normal. Postpartum testing distinguishes normal glucose tolerance from persistent prediabetes or diabetes; pregnancy may have revealed an underlying metabolic disorder rather than produced a wholly temporary one. Persistent abnormalities therefore require reclassification rather than continued use of the gestational diagnosis alone. (niddk.nih.gov)

The American Diabetes Association’s 2026 Standards of Care specify a fasting, 75-g oral glucose tolerance test (OGTT) at 4–12 weeks postpartum, interpreted using nonpregnancy diagnostic criteria. The test measures fasting and two-hour plasma glucose and can identify impaired glucose tolerance missed by fasting measurement alone. Early postpartum glycated hemoglobin (HbA1c) is less suitable: increased red-cell turnover during pregnancy and blood loss at delivery may lower it, while its approximately three-month averaging period still includes glucose exposure during pregnancy. Thus, an apparently reassuring HbA1c may not exclude postpartum dysglycemia. (diabetesjournals.org)

Recurrence in a subsequent pregnancy and progression to later type 2 diabetes are distinct outcomes. A population-based study of Massachusetts births in 1998–2007, published in 2015, estimated recurrence at 34–48%, depending on how diabetes was identified and categorized in linked records. These percentages concern women with another pregnancy, not all women previously diagnosed with GDM. Later diabetes risk also varies with population, diagnostic criteria, and follow-up duration. In the HAPO Follow-up Study, over half of mothers whose pregnancy glucose retrospectively met newer GDM criteria had diabetes or prediabetes 10–14 years later, compared with 20.1% of mothers below those criteria. This combined endpoint must not be interpreted as the incidence of diabetes alone. (stacks.cdc.gov)

Offspring follow-up has identified associations with both adiposity and glucose metabolism. HAPO follow-up included 4,832 children aged 10–14 years: maternal GDM defined retrospectively was associated with childhood obesity, although the broader combined outcome of overweight or obesity was not significantly different between groups after adjustment. Higher maternal pregnancy glucose was also associated with greater childhood adiposity, lower insulin sensitivity, and impaired glucose metabolism. These are observational findings, not proof that intrauterine glucose exposure alone caused the later abnormalities. Shared genetic susceptibility, maternal adiposity, and the family’s postnatal environment complicate separation of prenatal effects from familial influences; statistical adjustment cannot eliminate all confounding. (niddk.nih.gov)

Evidence concerning lactation is likewise primarily observational. In Erica Gunderson and colleagues’ 2015 SWIFT cohort, 959 women without diabetes at the initial postpartum assessment were evaluated during two years of follow-up; 113 (11.8%) developed diabetes. Exclusive breastfeeding, compared with exclusive formula feeding, was associated with an adjusted hazard ratio of 0.46, and longer lactation showed a graded association with lower incidence. The prospective design and repeated OGTTs strengthened ascertainment, but feeding was not randomized, so residual differences between groups remain possible explanations for part of the association. (pmc.ncbi.nlm.nih.gov)

Randomized prevention evidence concerns selected women with elevated metabolic risk, rather than every woman immediately after GDM. In Vanita Aroda and colleagues’ 2015 report from the Diabetes Prevention Program and its follow-up, intensive lifestyle intervention and metformin reduced progression to diabetes over ten years by approximately 35% and 40%, respectively, compared with placebo among participants with previous GDM. Participants entered with elevated glucose and increased body mass; the results demonstrate prevention or delay in that high-risk group. For ongoing surveillance, the ADA’s 2026 guidance specifies testing for prediabetes or type 2 diabetes every 1–3 years, even after a normal postpartum OGTT. (pubmed.ncbi.nlm.nih.gov)

Development of the Evidence Base

Early diagnostic criteria emphasized the mother’s subsequent risk of diabetes mellitus, rather than a glucose level at which pregnancy complications began. In 1964, John B. O’Sullivan and Claire M. Mahan published criteria based on a three-hour, 100-g oral glucose tolerance test and validated them against later maternal diabetes. The National Diabetes Data Group’s 1979 revision and Marshall Carpenter and Donald Coustan’s 1982 revision adapted this framework to changes in laboratory measurement, including the transition from whole blood to plasma and the use of more glucose-specific assays. The later development of outcome-based criteria changed the central question from predicting future maternal disease to identifying hyperglycemia associated with adverse pregnancy outcomes. (pubmed.ncbi.nlm.nih.gov)

The Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study, published in 2008 by a collaborative group including Boyd E. Metzger, supplied large-scale observational evidence for this change. It enrolled 25,505 women at 15 centers in nine countries, with 75-g oral glucose tolerance testing at 24–32 weeks. The principal analysis included 23,316 pregnancies in which glucose results remained concealed from participants and clinicians under prespecified safety limits, reducing the influence of glucose-directed treatment on observed outcomes. Higher fasting, one-hour, and two-hour glucose concentrations showed graded associations with birth weight above the 90th percentile and elevated cord-blood C-peptide, a marker of fetal insulin secretion. For each standard-deviation increase in these three glucose measurements, the adjusted odds ratios for birth weight above the 90th percentile were 1.38, 1.46, and 1.38, respectively. Associations with first cesarean delivery and neonatal hypoglycemia were weaker. No clear biological threshold separated low-risk from high-risk pregnancies. (pubmed.ncbi.nlm.nih.gov)

Evidence that treatment could alter outcomes came from randomized controlled trials. In 2009, Mark B. Landon and colleagues in the Maternal–Fetal Medicine Units Network reported a trial involving 958 women at 24–31 weeks with mild gestational diabetes and fasting glucose below 95 mg/dL. Treatment comprised dietary intervention, glucose self-monitoring, and insulin when necessary; the comparator was usual prenatal care. The primary composite—stillbirth or perinatal death, hyperbilirubinemia, hypoglycemia, hyperinsulinemia, or birth trauma—did not differ significantly: 32.4% with treatment versus 37.0% with usual care ($P=0.14$). Several prespecified secondary outcomes nevertheless improved. Treatment reduced large-for-gestational-age births from 14.5% to 7.1%, births above 4,000 g from 14.3% to 5.9%, shoulder dystocia from 4.0% to 1.5%, and cesarean delivery from 33.8% to 26.9%. Combined preeclampsia and gestational hypertension rates fell from 13.6% to 8.6%; mean birth weight and neonatal fat mass also decreased. (pubmed.ncbi.nlm.nih.gov)

In 2010, the International Association of Diabetes and Pregnancy Study Groups (IADPSG) translated HAPO’s continuous risk relationships into consensus diagnostic recommendations. Its panel selected glucose concentrations corresponding to 1.75-fold odds of specified neonatal outcomes relative to the odds at mean glucose concentrations: birth weight, cord C-peptide, and neonatal body-fat percentage above the 90th percentile. This was a consensus choice along a risk continuum, not the discovery of a natural disease boundary. The World Health Organization adopted the same outcome-based principles in its 2013 classification, distinguishing gestational diabetes from diabetes mellitus first detected during pregnancy. These documents provided an international standardization framework, although the WHO graded its recommendation on gestational-diabetes diagnostic cutoffs as weak, with very-low-quality evidence. (pmc.ncbi.nlm.nih.gov)

HAPO follow-up extended investigation to metabolic health 10–14 years after delivery. A 2018 analysis by William L. Lowe Jr. and colleagues included 4,697 mothers and 4,832 children, with gestational diabetes classified retrospectively using IADPSG criteria. Among mothers with that classification, 52.2% had a glucose-metabolism disorder at follow-up, compared with 20.1% without it. Childhood obesity and several direct measures of adiposity were associated with maternal gestational diabetes, whereas the adjusted difference in the combined overweight-or-obesity outcome was not statistically significant. Subsequent analyses published in 2019 examined glucose metabolism in 4,160 children, including glucose tolerance, insulin sensitivity, and secretion. This follow-up design investigated long-term associations; it did not establish that intrauterine glucose exposure, rather than shared familial influences, caused the later outcomes. (pubmed.ncbi.nlm.nih.gov)

Prevention Research and Health-System Implementation

Antenatal prevention trials examine whether changes in nutrition, physical activity, and gestational weight gain reduce GDM incidence before treatment becomes necessary. Results differ across populations and intervention designs. In the Finnish RADIEL randomized controlled trial, published in 2016, Saila B. Koivusalo and colleagues enrolled 293 women before 20 weeks of gestation who had previous GDM, a prepregnancy body mass index of at least 30 kg/m², or both. Individualized counseling by nurses, supplemented by a dietitian-led group session, produced GDM rates of 13.9%, compared with 21.6% under usual care among the 269 participants analyzed; the adjusted comparison was statistically significant. By contrast, the 2015 UK UPBEAT trial recruited 1,555 pregnant women with obesity and found no significant reduction: GDM occurred in approximately 25% of the intervention group and 26% of controls. These findings do not establish a universally effective prevention program; they demonstrate that a successful intervention in one selected population may not reproduce its effect elsewhere. The American Diabetes Association’s 2026 evidence synthesis identifies first-trimester or early-second-trimester initiation as a potentially important feature, while acknowledging heterogeneity. (pubmed.ncbi.nlm.nih.gov)

Implementation requires a connected health-system pathway rather than an isolated diagnostic test. Screening generates laboratory appointments, glucose-load administration, timed blood sampling, result interpretation, and communication with maternity services. A two-step pathway can require a second attendance for diagnostic testing, whereas a one-step pathway requires every tested participant to undergo the longer diagnostic procedure. The NIDDK’s Diabetes in America chapter describes a Canadian randomized economic comparison that included participants’ time costs: two-step testing was modestly less expensive overall, but this difference was absent in the Asian subgroup, where GDM prevalence was higher. After diagnosis, service requirements include glucose meters and testing supplies, education, nutrition services, medication access when indicated, and coordination between obstetric and diabetes professionals. Consequently, testing costs alone do not represent the full resource requirements of a screening policy. (niddk.nih.gov)

Continuity into primary health care is a separate implementation problem. In a 2018 NIDDK interview, Boyd E. Metzger identified competing childcare demands, fragmented records, and loss of awareness of long-term risk after glucose levels normalize as obstacles to postpartum follow-up. Reminder systems have measurable, but incomplete, effects. Heather D. Clark and colleagues’ 2009 Canadian randomized trial assessed completion of an oral glucose tolerance test within one year after delivery: completion reached 60.5% when reminders went to both patient and physician, 55.3% with patient-only reminders, and 51.6% with physician-only reminders, compared with 14.3% without reminders. Even the most effective arm left nearly two-fifths untested. The outcome was testing within one year, not compliance with the narrower early-postpartum testing window, an important distinction when evaluating service performance. (niddk.nih.gov)

Early-pregnancy hyperglycemia research concerns treatment before conventional screening, not prevention of incident GDM. David Simmons and colleagues’ 2023 TOBOGM trial randomized 802 women with risk factors and GDM diagnosed before 20 weeks to immediate treatment or deferred management based on repeat testing at 24–28 weeks. The adverse neonatal composite occurred in 24.9% versus 30.5%, respectively, but pregnancy-related hypertension did not differ materially. The 2026 ADA assessment nevertheless identifies unresolved questions about which early glucose abnormalities persist and which subgroups benefit from treatment. Monitoring research presents a similar distinction between demonstrated intermediate effects and clinical benefit: GDM-specific trials of continuous glucose monitoring have reported improvements in some glucose metrics, but one trial found no significant improvement in its primary perinatal and neonatal composite. Optimal GDM-specific sensor targets and evidence supporting universal use remain incomplete. (nejm.org)

Longer-term offspring research tests whether improved pregnancy outcomes translate into reduced childhood metabolic risk. A follow-up of the Maternal–Fetal Medicine Units Network mild-GDM treatment trial examined 500 children aged 5–10 years, representing 55% of eligible offspring. Obesity, defined by body mass index at or above the 95th percentile, occurred in 20.8% of the treated group and 22.9% of the untreated group, without a statistically significant difference. Incomplete participation and limited statistical power constrained detection of modest effects. Thus, neonatal treatment benefits cannot be assumed to establish prevention of childhood obesity; sustained follow-up of randomized cohorts addresses a question distinct from observational associations between maternal glucose and offspring health. (pubmed.ncbi.nlm.nih.gov)

References

1. Chapter 4: Gestational Diabetesniddk.nih.gov
2. Executive summary — Diagnostic Criteria and Classification of Hyperglycaemia First Detected in Pregnancyncbi.nlm.nih.gov
3. The timing of initiation of pharmacotherapy for women with gestational diabetes mellituspmc.ncbi.nlm.nih.gov
4. Gestational Diabetes - PubMedpubmed.ncbi.nlm.nih.gov
5. QuickStats: Percentage of Mothers with Gestational Diabetes, by Maternal Age — National Vital Statistics System, United States, 2016 and 2021cdc.gov
6. Births: Final Data for 2021stacks.cdc.gov
7. Symptoms & Causes of Gestational Diabetes - NIDDKniddk.nih.gov
8. Preventing Gestational Diabetes - NIDDKniddk.nih.gov
9. Gestational Diabetes | ACOGacog.org
10. Longitudinal changes in insulin release and insulin resistance in nonobese pregnant womenpubmed.ncbi.nlm.nih.gov
11. Human placental growth hormone causes severe insulin resistance in transgenic micepubmed.ncbi.nlm.nih.gov
12. Carbohydrate metabolism during pregnancy in control subjects and women with gestational diabetespubmed.ncbi.nlm.nih.gov
13. Heterogeneous Contribution of Insulin Sensitivity and Secretion Defects to Gestational Diabetes Mellituspmc.ncbi.nlm.nih.gov
14. Hyperglycemia and Adverse Pregnancy Outcome (HAPO) Study: Associations With Neonatal Anthropometricspmc.ncbi.nlm.nih.gov
15. Gestational diabetes - NHSnhs.uk
16. Recommendation: Gestational Diabetes: Screeninguspreventiveservicestaskforce.org
17. Recommendations: Diabetes in pregnancy: management from preconception to the postnatal periodnice.org.uk
18. Diabetes in pregnancy: management from preconception to the postnatal periodnice.org.uk
19. Tests & Diagnosis for Gestational Diabetes - NIDDKniddk.nih.gov
20. Recommendations — Diagnostic Criteria and Classification of Hyperglycaemia First Detected in Pregnancyncbi.nlm.nih.gov
21. Diagnostic Criteria and Classification of Hyperglycaemia First Detected in Pregnancyncbi.nlm.nih.gov
22. Diagnosis and Classification of Diabetes: Standards of Care in Diabetes—2026pmc.ncbi.nlm.nih.gov
23. Management of Diabetes in Pregnancy: Standards of Care in Diabetes—2026pmc.ncbi.nlm.nih.gov
24. Preeclampsianichd.nih.gov
25. Neonatal Complications: Recognition and Prompt Treatment of Shoulder Dystociapatientsafety.pa.gov
26. Glycaemic control in labour with diabetes: GILD, a scoping studyncbi.nlm.nih.gov
27. Am I at risk for gestational diabetes?nichd.nih.gov
28. Pregnancy if You Have Diabetesniddk.nih.gov
29. Managing & Treating Gestational Diabetesniddk.nih.gov
30. Effectiveness of telemedicine for pregnant women with gestational diabetes mellitus: an updated meta-analysis of 32 randomized controlled trials with trial sequential analysispmc.ncbi.nlm.nih.gov
31. SMFM Statement: Pharmacological treatment of gestational diabetess3.amazonaws.com
32. Society for Maternal-Fetal Medicine Statement: Pharmacological treatment of gestational diabetespublications.smfm.org
33. Recommendations: Diabetes in pregnancy—management from preconception to the postnatal periodnice.org.uk
34. Indications for Outpatient Antenatal Fetal Surveillanceacog.org
35. Estimating the Recurrence Rate of Gestational Diabetes Mellitus (GDM) in Massachusetts 1998–2007: Methods and Findingsstacks.cdc.gov
36. Pregnancy, metabolism, and the short- and long-term health of women and their childrenniddk.nih.gov
37. Hyperglycemia and Adverse Pregnancy Outcome Follow-Up Study: newborn anthropometrics and childhood glucose metabolismpmc.ncbi.nlm.nih.gov
38. Gestational diabetes mellitus and developmental programmingpmc.ncbi.nlm.nih.gov
39. Lactation and Progression to Type 2 Diabetes Mellitus After Gestational Diabetes Mellitus: A Prospective Cohort Studypmc.ncbi.nlm.nih.gov
40. The effect of lifestyle intervention and metformin on preventing or delaying diabetes among women with and without gestational diabetes: the Diabetes Prevention Program outcomes study 10-year follow-uppubmed.ncbi.nlm.nih.gov
41. Criteria for the Oral Glucose Tolerance Test in Pregnancypubmed.ncbi.nlm.nih.gov
42. Background — Diagnostic Criteria and Classification of Hyperglycaemia First Detected in Pregnancyncbi.nlm.nih.gov
43. International Association of Diabetes and Pregnancy Study Groups Recommendations on the Diagnosis and Classification of Hyperglycemia in Pregnancypmc.ncbi.nlm.nih.gov
44. Hyperglycemia and Adverse Pregnancy Outcomespubmed.ncbi.nlm.nih.gov
45. A Multicenter, Randomized Trial of Treatment for Mild Gestational Diabetespubmed.ncbi.nlm.nih.gov
46. Association of Gestational Diabetes With Maternal Disorders of Glucose Metabolism and Childhood Adipositypubmed.ncbi.nlm.nih.gov
47. Hyperglycemia and Adverse Pregnancy Outcome Follow-up Study (HAPO FUS): Maternal Glycemia and Childhood Glucose Metabolismpubmed.ncbi.nlm.nih.gov
48. Hyperglycemia and Adverse Pregnancy Outcome Follow-up Study (HAPO FUS): Maternal Gestational Diabetes Mellitus and Childhood Glucose Metabolismteamabc3.wordpress.com
49. Gestational Diabetes Mellitus Can Be Prevented by Lifestyle Intervention: The Finnish Gestational Diabetes Prevention Study (RADIEL): A Randomized Controlled Trialpubmed.ncbi.nlm.nih.gov
50. Effect of a behavioural intervention in obese pregnant women (the UPBEAT study): a multicentre, randomised controlled trialeprints.gla.ac.uk
51. Providing Care for Women with a History of Gestational Diabetes: What’s Your Role?niddk.nih.gov
52. Do postal reminders increase postpartum screening of diabetes mellitus in women with gestational diabetes mellitus? A randomized controlled trialpubmed.ncbi.nlm.nih.gov
53. Mild gestational diabetes mellitus and long-term child healthpubmed.ncbi.nlm.nih.gov
54. Mild Gestational Diabetes Mellitus and Long-Term Child Healthpmc.ncbi.nlm.nih.gov