POSITION STATEMENT

Gestational Diabetes Mellitus

AMERICAN DIABETES ASSOCIATION

DEFINITION, DETECTION, ● Member of an ethnic group with a low Mahan, modified by Carpenter and Cous- AND DIAGNOSIS prevalence of GDM tan, and are shown in Table 1. Alterna- ● No known diabetes in first-degree relatives tively, the diagnosis can be made using a Definition ● No history of abnormal glucose tolerance 75-g glucose load and the glucose thresh- Gestational diabetes mellitus (GDM) is ● No history of poor obstetric outcome old values listed for fasting, 1 h, and 2 h defined as any degree of glucose intoler- (Table 2); however, this test is not as well ance with onset or first recognition during A fasting plasma glucose level Ͼ126 validated for detection of at-risk infants or pregnancy (1). The definition applies mg/dl (7.0 mmol/l) or a casual plasma mothers as the 100-g OGTT. whether insulin or only diet modification glucose Ͼ200 mg/dl (11.1 mmol/l) meets is used for treatment and whether or not the threshold for the diagnosis of diabe- OBSTETRIC AND PERINATAL the condition persists after pregnancy. It tes, if confirmed on a subsequent day, and CONSIDERATIONS — The pres- does not exclude the possibility that un- precludes the need for any glucose chal- ence of fasting hyperglycemia (Ͼ105 recognized glucose intolerance may have lenge. In the absence of this degree of hy- mg/dl or Ͼ5.8 mmol/l) may be associated antedated or begun concomitantly with perglycemia, evaluation for GDM in with an increase in the risk of intrauterine the pregnancy. women with average or high-risk charac- fetal death during the last 4–8 weeks of Approximately 7% of all pregnancies teristics should follow one of two ap- gestation. Although uncomplicated GDM are complicated by GDM, resulting in proaches: with less severe fasting hyperglycemia has more than 200,000 cases annually. The not been associated with increased peri- prevalence may range from 1 to 14% of all One-step approach: Perform a diagnos- natal mortality, GDM of any severity in- pregnancies, depending on the popula- tic oral glucose tolerance test (OGTT) creases the risk of fetal macrosomia. tion studied and the diagnostic tests em- without prior plasma or serum glucose Neonatal hypoglycemia, jaundice, poly- ployed. screening. The one-step approach may be cythemia, and hypocalcemia may compli- cost-effective in high-risk patients or pop- cate GDM as well. GDM is associated with ulations (e.g., some Native-American Detection and diagnosis an increased frequency of maternal hy- groups). Risk assessment for GDM should be un- pertensive disorders and the need for ce- sarean delivery. The latter complication dertaken at the first prenatal visit. Women Two-step approach: Perform an initial with clinical characteristics consistent may result from fetal growth disorders screening by measuring the plasma or se- and/or alterations in obstetric manage- with a high risk of GDM (marked obesity, rum glucose concentration 1 h after a personal history of GDM, glycosuria, or a ment due to the knowledge that the 50-g oral glucose load (glucose challenge mother has GDM. strong family history of diabetes) should test [GCT]) and perform a diagnostic undergo glucose testing (see below) as OGTT on that subset of women exceeding soon as feasible. If they are found not to the glucose threshold value on the GCT. Long-term considerations have GDM at that initial screening, they When the two-step approach is em- Women with GDM are at increased risk should be retested between 24 and 28 ployed, a glucose threshold value Ͼ140 for the development of diabetes, usually weeks of gestation. Women of average mg/dl (7.8 mmol/l) identifies approxi- type 2, after pregnancy. Obesity and other risk should have testing undertaken at mately 80% of women with GDM, and the factors that promote insulin resistance ap- 24–28 weeks of gestation. Low-risk sta- yield is further increased to 90% by using pear to enhance the risk of type 2 diabetes tus requires no glucose testing, but this a cutoff of Ͼ130 mg/dl (7.2 mmol/l). after GDM, while markers of islet cell– category is limited to those women meet- directed autoimmunity are associated ing all of the following characteristics: With either approach, the diagnosis with an increase in the risk of type 1 dia- of GDM is based on an OGTT. Diagnostic betes. Offspring of women with GDM are ● Age Ͻ25 years criteria for the 100-g OGTT are derived at increased risk of obesity, glucose intol- ● Weight normal before pregnancy from the original work of O’Sullivan and erance, and diabetes in late adolescence ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● and young adulthood. The recommendations in this paper are based on the evidence reviewed in the following publications: Report of the expert committee on the diagnosis and classification of diabetes mellitus. Diabetes Care 21 (Suppl. THERAPEUTIC STRATEGIES 1):S5–S19, 1998; and the Proceedings of the 4th International Workshop-Conference on Gestational Dia- DURING PREGNANCY betes Mellitus. Diabetes Care 21 (Suppl. 2):B1–B167, 1998. Originally approved 1986. Most recent review/revision, 2000. Abbreviations: GCT, glucose challenge test; GDM, gestational diabetes mellitus; IFG, impaired fasting Monitoring ● glucose; IGT, impaired glucose tolerance; MNT, medical nutrition therapy; OGTT, oral glucose tolerance Maternal metabolic surveillance should test; SMBG, self-monitoring of blood glucose. be directed at detecting hyperglycemia

DIABETES CARE, VOLUME 26, SUPPLEMENT 1, JANUARY 2003 S103 Position Statement

Table 1—Diagnosis of GDM with a 100-g Table 2—Diagnosis of GDM with a 75-g ommended when MNT fails to main- oral glucose load oral glucose load tain self-monitored glucose at the following levels: mg/dl mmol/l mg/dl mmol/l Fasting whole blood glucose Յ Fasting 95 5.3 Fasting 95 5.3 95 mg/dl (5.3 mmol/l) Fasting plasma glucose 1-h 180 10.0 1-h 180 10.0 Յ 2-h 155 8.6 2-h 155 8.6 105 mg/dl (5.8 mmol/l) or 3-h 140 7.8 Two or more of the venous plasma concentrations must be met or exceeded for a positive diagnosis. 1-h postprandial whole blood glucose Two or more of the venous plasma concentrations Յ must be met or exceeded for a positive diagnosis. The test should be done in the morning after an 140 mg/dl (7.8 mmol/l) The test should be done in the morning after an overnight fast of between 8 and 14 h and after at least 1-h postprandial plasma glucose Ն overnight fast of between 8 and 14 h and after at least 3 days of unrestricted diet ( 150 g carbohydrate Յ155 mg/dl (8.6 mmol/l) Ն per day) and unlimited physical activity. The subject 3 days of unrestricted diet ( 150 g carbohydrate per or day) and unlimited physical activity. The subject should remain seated and should not smoke should remain seated and should not smoke throughout the test. 2-h postprandial whole blood glucose throughout the test. Յ120 mg/dl (6.7 mmol/l) 2-h postprandial plasma glucose pending on maternal weight and height Յ130 mg/dl (7.2 mmol/l) severe enough to increase risks to the is recommended. MNT should include fetus. Daily self-monitoring of blood the provision of adequate calories and ● Measurement of the fetal abdominal glucose (SMBG) appears to be superior nutrients to meet the needs of preg- circumference early in the third trimes- to intermittent office monitoring of nancy and should be consistent with ter can identify a large subset of infants plasma glucose. For women treated the maternal blood glucose goals that with no excess risk of macrosomia in with insulin, limited evidence indicates have been established. Noncaloric the absence of maternal insulin ther- that postprandial monitoring is supe- sweeteners may be used in moderation. apy. This approach has been tested pri- rior to preprandial monitoring. How- ● For obese women (BMI Ͼ30 kg/m2), a marily in pregnancies with maternal ever, the success of either approach 30–33% calorie restriction (to ϳ25 fasting serum glucose levels Ͻ105 depends on the glycemic targets that kcal/kg actual weight per day) has been mg/dl (5.8 mmol/l). are set and achieved. shown to reduce hyperglycemia and ● Human insulin should be used when ● Urine glucose monitoring is not useful plasma triglycerides with no increase in insulin is prescribed, and SMBG should in GDM. Urine ketone monitoring may ketonuria (2). Restriction of carbohy- guide the doses and timing of the insu- be useful in detecting insufficient ca- drates to 35–40% of calories has been lin regimen. The use of insulin analogs loric or carbohydrate intake in women shown to decrease maternal glucose has not been adequately tested in GDM. treated with calorie restriction. levels and improve maternal and fetal ● Oral glucose-lowering agents have gen- ● Maternal surveillance should include outcomes (3). erally not been recommended during blood pressure and urine protein mon- ● Insulin is the pharmacologic therapy pregnancy. However, one randomized, itoring to detect hypertensive disor- that has most consistently been shown unblinded clinical trial compared the ders. to reduce fetal morbidities when added use of insulin and glyburide in women ● Increased surveillance for pregnancies to MNT. Selection of pregnancies for with GDM who were not able to meet at risk for fetal demise is appropriate, insulin therapy can be based on mea- glycemic goals on MNT (4). Treatment particularly when fasting glucose lev- sures of maternal glycemia with or with either agent resulted in similar els exceed 105 mg/dl (5.8 mmol/l) or without assessment of fetal growth perinatal outcomes. All patients were pregnancy progresses past term. The characteristics. When maternal glucose beyond the first trimester of pregnancy initiation, frequency, and specific tech- levels are used, insulin therapy is rec- at the initiation of therapy. Glyburide is niques used to assess fetal well-being will depend on the cumulative risk the fetus bears from GDM and any other Table 3—Criteria for the diagnosis of diabetes mellitus medical/obstetric conditions present. ● Assessment for asymmetric fetal growth Normoglycemia IFG and IGT Diabetes mellitus* by ultrasonography, particularly in Ͻ Ն Ն early third trimester, may aid in identi- FPG 110 mg/dl FPG 110 mg/dl and FPG 126 mg/dl Ͻ126 mg/dl (IFG) fying fetuses that can benefit from ma- Ͻ Ն Ն ternal insulin therapy (see below). 2-h PG† 140 mg/dl 2-h PG† 140 mg/dl and 2-h PG† 200 mg/dl Ͻ200 mg/dl (IGT) Management ——Symptoms of DM and casual ● plasma glucose All women with GDM should receive Ն nutritional counseling, by a registered concentration 200 mg/dl dietitian when possible, consistent with DM, diabetes mellitus; FPG, fasting plasma glucose; 2-h PG, 2-h postload glucose. *A diagnosis of diabetes must be confirmed on a subsequent day by any one of the three methods included in the chart. In clinical the recommendations by the American settings, the FPG test is greatly preferred because of ease of administration, convenience, acceptability to Diabetes Association. Individualization patients, and lower cost. Fasting is defined as no calorie intake for at least 8 h. †This test requires the use of of medical nutrition therapy (MNT) de- a glucose load containing the equivalent of 75 g anhydrous glucose dissolved in water.

S104 DIABETES CARE, VOLUME 26, SUPPLEMENT 1, JANUARY 2003 Gestational Diabetes Mellitus

not FDA approved for the treatment of and according to the guidelines of the Offspring of women with GDM GDM and further studies are needed in “Report of the Expert Committee on the should be followed closely for the devel- a larger patient population to establish Diagnosis and Classification of Diabetes opment of obesity and/or abnormalities of its safety. Mellitus” (5). See Table 3 for diagnostic glucose tolerance. ● Programs of moderate physical exercise criteria. If glucose levels are normal post- have been shown to lower maternal partum, reassessment of glycemia should glucose concentrations in women with be undertaken at a minimum of 3-year GDM. Although the impact of exercise intervals. Women with IFG or IGT in the References on neonatal complications awaits rigor- postpartum period should be tested for 1. Metzger BE, Coustan DR (Eds.): Proceed- ous clinical trials, the beneficial glucose- diabetes annually; these patients should ings of the Fourth International Work- lowering effects warrant a recommen- receive intensive MNT and should be shop-Conference on Gestational Diabetes dation that women without medical or placed on an individualized exercise pro- Mellitus. Diabetes Care 21 (Suppl. 2):B1– B167, 1998 obstetrical contraindications be en- gram because of their very high risk for 2. Franz MJ, Horton ES, Bantle JP, Beebe CA, couraged to start or continue a program development of diabetes. All patients with Brunzell JD, Coulston AM, Henry RR, of moderate exercise as a part of treat- prior GDM should be educated regarding Hoogwerf BJ, Stacpoole PW: Nutrition ment for GDM. lifestyle modifications that lessen insulin principles for the management of diabetes ● GDM is not of itself an indication for resistance, including maintenance of nor- and related complications (Technical Re- cesarean delivery or for delivery before mal body weight through MNT and phys- view). Diabetes Care 17:490–518, 1994 38 completed weeks of gestation. Pro- ical activity. Medications that worsen 3. Major CA, Henry MJ, De Veciana M, Mor- longation of gestation past 38 weeks in- insulin resistance (e.g., glucocorticoids, gan MA: The effects of carbohydrate re- creases the risk of fetal macrosomia nicotinic acid) should be avoided if pos- striction in patients with diet-controlled without reducing cesarean rates, so that sible. Patients should be advised to seek gestational diabetes. Obstet Gynecol 91: delivery during the 38th week is recom- medical attention if they develop symp- 600–604, 1998 mended unless obstetric considerations toms suggestive of hyperglycemia. Educa- 4. Langer L, Conway DL, Berkus MD, Xenakis dictate otherwise. tion should also include the need for EM-J, Gonzales O: A comparison of gly- ● buride and insulin in women with gesta- Breast-feeding, as always, should be en- family planning to ensure optimal glyce- tional diabetes mellitus. N Engl J Med 343: couraged in women with GDM. mic regulation from the start of any sub- 1134–1138, 2000 sequent pregnancy. Low-dose estrogen- 5. Expert Committee on the Diagnosis and LONG-TERM THERAPEUTIC progestogen oral contraceptives may be Classification of Diabetes Mellitus: Report CONSIDERATIONS — Reclassifica- used in women with prior histories of of the Expert Committee on the Diagnosis tion of maternal glycemic status should be GDM, as long as no medical contraindi- and Classification of Diabetes Mellitus. performed at least 6 weeks after delivery cations exist. Diabetes Care 26 (Suppl. 1):S5–S20, 2003

DIABETES CARE, VOLUME 26, SUPPLEMENT 1, JANUARY 2003 S105