FEATURE ARTICLE

Gestational Diabetes Mellitus

Tracy L. Setji, MD; Ann J. Brown, MD; and Mark N. Feinglos, MD, CM

estational diabetes mellitus Table 1. ADA and WHO Criteria for the Diagnosis of GDM2,6 (GDM) is defined as glucose Gintolerance that begins or is first ADA ADA WHO detected during .1–3 GDM 100-g OGTT 75-g OGTT 75-g OGTT affects ~ 7% of all , resulting Fasting (mg/dl) 95 95 126 in > 200,000 cases per year.2 Depending 1-hour (mg/dl) 180 180 — on the population sample and diagnostic 2-hour (mg/dl) 155 155 140 criteria, the prevalence may range from 3-hour (mg/dl) 140 — — 1 to 14%.1,2 Of all pregnancies compli- For the ADA criteria, two or more of the values from either the 100- or 75-g OGTT must cated by diabetes, GDM accounts for 1 be met or exceeded to make the diagnosis of GDM. For the WHO criteria, one of the two ~ 90%. values from the 75-g OGTT must be met or exceeded to make the diagnosis of GDM.

DIAGNOSTIC CRITERIA rizes ADA and WHO criteria for the mia by the ADA criteria or for perinatal The oral glucose tolerance test (OGTT) diagnosis of GDM. death by the WHO criteria. This study most commonly used to diagnose GDM The Brazilian Gestational Diabetes concluded that, although the WHO crite- in the United States is the 3-hour, 100-g Study evaluated the ADA and WHO ria identified more cases of GDM, both OGTT. According to diagnostic criteria diagnostic criteria against pregnancy the ADA and WHO criteria are valid recommended by the American Diabetes outcomes in an observational cohort options for the diagnosis of GDM and Association (ADA), GDM is diagnosed study of nearly 5,000 women. Using the the prediction of adverse pregnancy out- if two or more plasma glucose levels 2-hour 75-g OGTT criteria proposed by comes.6 meet or exceed the following thresholds: the ADA, the incidence of GDM was fasting glucose concentration of 95 2.4% (95% CI 2.0–2.9). The incidence PATHOGENESIS mg/dl, 1-hour glucose concentration of of GDM using the WHO criteria was Pregnancy is a diabetogenic condition 180 mg/dl, 2-hour glucose concentration 7.2% (6.5–7.9). Both the ADA and characterized by insulin resistance with of 155 mg/dl, or 3-hour glucose concen- WHO criteria predicted an increased risk a compensatory increase in -cell tration of 140 mg/dl.1,2,4 These values are of macrosomia, preeclampsia, and peri- response and hyperinsulinemia. Insulin lower than the thresholds recommended natal death. However, this increase was resistance usually begins in the second by the National Diabetes Data Group not statistically significant for macroso- trimester and progresses throughout the and are based on the Carpenter and remainder of the pregnancy. Insulin sen- Coustan modification.5 The ADA rec- IN BRIEF sitivity is reduced by as much as 80%. ommendations also include the use of a Placental secretion of hormones, such as 2-hour 75-g OGTT with the same glu- Gestational diabetes mellitus (GDM) progesterone, cortisol, placental lacto- cose thresholds listed for fasting, 1-hour, is a common condition affecting gen, prolactin, and growth hormone, is a and 2-hour values.1,2 ~ 7% of all pregnancies. The detec- major contributor to the insulin-resistant The World Health Organization tion of GDM is important because of state seen in pregnancy. The insulin (WHO) diagnostic criteria, which are its associated maternal and fetal com- resistance likely plays a role in ensuring used in many countries outside of plications. Treatment with medical that the fetus has an adequate supply of North America, are based on a 2-hour nutrition therapy, close monitoring of glucose by changing the maternal ener- 75-g OGTT. GDM is diagnosed by glucose levels, and insulin therapy if gy metabolism from carbohydrates to WHO criteria if either the fasting glu- glucose levels are above goal can lipids.7 cose is > 126 mg/dl or the 2-hour glu- help to reduce these complications. Women with GDM have a greater cose is > 140 mg/dl. Table 1 summa- severity of insulin resistance compared

CLINICAL DIABETES • Volume 23, Number 1, 2005 17 FEATURE ARTICLE

to the insulin resistance seen in normal Neonatal hypoglycemia can occur However, Naylor et al.22 developed a pregnancies. They also have an impair- within a few hours of delivery. This selective screening approach with data ment of the compensatory increase in results from maternal hyperglycemia collected from 3,131 pregnant women. insulin secretion, particularly first-phase causing fetal hyperinsulinemia.7 They randomly selected data from half insulin secretion. This decrease in first- The association between GDM and of the women and categorized them into phase insulin release may be a marker perinatal mortality has been more con- three groups (low-, intermediate-, and for deterioration of -cell function.7 troversial. Several studies have conclud- high-risk) based on a complex scoring Xiang et al.8 found that Latino women ed that the rate of perinatal mortality system using weighted risk factors: age, with GDM had increased resistance to (including stillbirths and neonatal BMI before pregnancy, and race. Their the effects of insulin on glucose clear- deaths) was increased in women with strategy did not entail screening women ance and production compared with nor- GDM in the past.14–16 However, recent who were low risk. The remainder of mal pregnant women. In addition, they studies have shown that, with the combi- women were screened with a 50-g oral found that the women with GDM had a nation of increased antepartum monitor- glucose challenge test, with the thresh- 67% reduction in their -cell compensa- ing, medical nutrition therapy (MNT), old for a positive result based on their tion compared with normal pregnant and insulin therapy if needed, this differ- risk score. Naylor et al. found that this control subjects. ence in perinatal mortality rates is poten- selective screening approach resulted in There is also a subset of women tially avoidable.17,18 a 34.6% reduction in the number of with GDM who have evidence of islet Another controversial association screening tests performed, without a cell autoimmunity. The reported preva- is that between GDM and congenital decrease in the detection rate of GDM.22 lence of islet cell antibodies in women malformations. The incidence of a The ADA now recommends selective with GDM ranges from 1.6 to 38%. The major malformation in an infant whose screening for GDM.1,2 prevalence of other islet autoantibodies, mother does not have any history of According to the ADA guidelines, including insulin autoantibodies and diabetes is 1–3%. In women with a his- patients should be screened for risk fac- glutamic acid decarboxylase antibodies, tory of diabetes before pregnancy, this tors for GDM at their initial visit. A has also been variable. These women risk is increased three to eight times.19 woman is considered high risk if she has may be at risk for developing an In women with GDM, an increased one or more of the following: marked autoimmune form of diabetes later in incidence of malformations occurs obesity, personal history of GDM, glu- life.9 Finally, in ~ 5% of all cases of when the mother also has fasting cose intolerance or glycosuria, or a GDM, the -cell’s inability to compen- hyperglycemia.19,20 strong family history of type 2 diabetes. sate for the insulin resistance is the Long-term complications to the off- A woman is considered low risk risk if result of a defect in the -cell, such as a spring include an increased risk of glu- she meets all of the following criteria: mutation in glucokinase.7 cose intolerance, diabetes, and obesity.2 age < 25 years, normal prepregnancy Maternal complications associated weight, not a member of an ethnic/racial COMPLICATIONS with GDM include hypertension, group with a high prevalence of diabetes There are both fetal and maternal com- preeclampsia, and an increased risk of (e.g., Hispanic American, Native Ameri- plications associated with GDM. Fetal cesarean delivery.6,11 The hypertension can, Asian American, African American, complications include macrosomia, may be related to insulin resistance. or Pacific Islander), no known diabetes neonatal hypoglycemia, perinatal mor- Therefore, interventions that improve in first-degree relatives, no history of tality, congenital malformation, hyper- insulin sensitivity may help prevent this abnormal glucose tolerance, and no his- bilirubinemia, polycythemia, hypocal- complication.21 In addition, women with tory of a poor obstetric outcome. A cemia, and respiratory distress syn- a history of GDM have an increased risk woman is considered intermediate risk if drome.1–3,10–19 Macrosomia, defined as of developing diabetes after pregnancy she does not fall into either the high- or > 4,000 g,7 occurs in compared to the general population, with low-risk category. ~ 20–30% of infants whose mothers a conversion rate of up to 3% per year.7 If a woman is high risk, glucose test- have GDM.10 Maternal factors associat- ing should be done as soon as possible. ed with an increased incidence of SCREENING If the initial testing is negative, the macrosomia include hyperglycemia,3,10,11 There is no worldwide agreement on the woman should be retested between 24 to high BMI,7,10–12 older age,11 and multi- best way to screen for GDM. Previously, 28 weeks of gestation. If she is interme- parity.7,11 This excess in fetal growth can universal screening at 24–28 weeks of diate risk, she should undergo glucose lead to increased fetal morbidity at gestation with a 50-g oral glucose chal- testing at 24 to 28 weeks. If she is low delivery, such as shoulder dystocia,11 lenge test was recommended. Women risk, the ADA does not recommend and an increased rate of cesarean with a 1-hour glucose level > 140 mg/dl screening for GDM.1–3 An additional deliveries.11,13 were referred for a diagnostic OGTT. possible risk factor for GDM not men-

18 Volume 23, Number 1, 2005 • CLINICAL DIABETES FEATURE ARTICLE

tioned in the list above is a history of rate (> 80%) of the 50-g glucose chal- therapy, adjustments of their insulin regi- polycystic ovary syndrome.23,24 However, lenge test.4 mens based on postprandial, rather than other studies have not confirmed this preprandial, glucose levels decreased the finding.25,26 TREATMENT incidence of neonatal hypoglycemia, The ADA recommends two macrosomia, and cesarean delivery for approaches to screening for GDM if a Glucose Monitoring cephalopelvic disproportion.31 woman has one or more risk factors—a Self-monitoring of blood glucose is one-step or a two-step approach. The recommended for women with GDM. Medical Nutrition Therapy more commonly used two-step approach The goal of monitoring is to detect The goals of MNT are to provide ade- involves initial nonfasting screening with glucose concentrations elevated quate nutrition for the mother and fetus, the 50-g oral glucose challenge test, fol- enough to increase perinatal mortality. provide sufficient calories for appropri- lowed by a 1-hour serum glucose con- The Fourth International Workshop- ate maternal weight gain, maintain nor- centration. If the glucose level exceeds Conference on Gestational Diabetes moglycemia, and avoid ketosis. In gen- the glucose threshold value on this test, Mellitus recommends maintaining the eral, there is not an increased energy the patient is further evaluated with the following capillary blood glucose val- requirement during the first trimester of diagnostic OGTT described previously ues: preprandial glucose < 95 mg/dl, 1- pregnancy. However, most normal- under diagnostic criteria. A 1-hour glu- hour postprandial glucose < 140 mg/dl, weight women require an additional 300 cose value > 140 mg/dl identifies ~ 80% and 2-hour postprandial glucose < 120 kcal/day in the second and third of women with GDM. A 1-hour glucose mg/dl.3 ACOG guidelines are the same trimester.32 value > 130 mg/dl identifies ~ 90% of except that the 1-hour postprandial In normal-weight women with women with GDM, but it has a higher glucose value is considered acceptable GDM, the recommended daily caloric false positive rate.1–3 Either value is at either 130 or 140 mg/dl.27 intake is 30 kcal/kg/day based on their accepted by the ADA and the American Jovanovic-Peterson et al.28 suggest present pregnant weight.33 In women College of Obstetricians and Gynecolo- guidelines that are a little stricter: fast- with GDM who are overweight (BMI > gists (ACOG) as abnormal.1,2,27 ing glucose < 90 mg/dl and 1-hour 30 kg/m2), a 33% calorie restriction of The one-step approach requires a postprandial glucose < 120 mg/dl.28 their estimated energy needs is recom- diagnostic OGTT without prior screen- One prospective study of 668 mended (~ 25 kcal/kg/day based on their ing with the 50-g 1-hour glucose chal- patients (334 with GDM and 334 control present pregnant weight). This level of lenge test. This may be cost-effective in subjects) found that women with GDM calorie restriction is not associated with some high-risk patients. Of note, if a who had a mean blood glucose level an elevation of free fatty acids or patient has a fasting plasma glucose level between 87 and 104 mg/dl had incidence ketonuria.32,34 Some authors recommend > 126 mg/dl or a random plasma glucose rates of intrauterine growth retardation further calorie restriction for women level > 200 mg/dl, this meets the thresh- (IUGR) and large for who are morbidly obese.33 However, old for diabetes mellitus and should be (LGA) infants comparable to the control caution must be taken to avoid ketosis, confirmed on a subsequent day.1–3 group. However, women who had mean which can be seen with more aggressive Screening, whether it is universal or blood glucose values < 87 mg/dl had a calorie restriction.34 selective, remains a controversial sub- higher incidence of infants with IUGR, Ketonemia in mothers with diabetes ject. Contradictory to the ADA recom- whereas women who had mean blood during pregnancy has been associated mendations described above, the United glucose values > 104 mg/dl had a higher with lower IQ levels and impaired psy- States Preventive Services Task Force incidence of LGA infants. This study chomotor development in their chil- concluded that there was insufficient evi- suggests that although it is important to dren.35,36 Monitoring with prebreakfast dence to recommend for or against treat hyperglycemia in GDM, it is also ketone measurements is recommended screening for GDM. Although they important not to overtreat because this for patients who are on a hypocaloric or found fair to good evidence that screen- can increase the risk of IUGR.29 carbohydrate-restricted diet.32 ing and treatment of GDM reduced the It is important for women to check Carbohydrates should be distributed rate of fetal macrosomia, they found postprandial glucose levels because throughout the day.32 Eating three small- insufficient evidence that screening sig- these have been shown to correlate more to moderate-sized meals and three nificantly reduced important adverse with macrosomia than do fasting levels. snacks per day is recommended. Limit- maternal or fetal outcomes, including The Diabetes in Early Pregnancy Study ing carbohydrates to 40% of the total outcomes related to macrosomia. In found that third-trimester postprandial daily caloric intake has been shown to addition, they had concerns about the glucose levels were the strongest predic- decrease postprandial glucose levels.33 potential harms and costs of screening, tor of percentile birth weight.30 In Further limitation of carbohydrates at especially given the high false-positive women with GDM who require insulin breakfast to 33% may be required to

CLINICAL DIABETES • Volume 23, Number 1, 2005 19 FEATURE ARTICLE

meet the desired postprandial glucose found that the diet-and-exercise group women in the conventional management goals because insulin resistance is great- had a significant decrease in glycated group. est in the morning.37 In addition, carbo- hemoglobin levels and in both fasting This study demonstrated a decreased hydrate restriction to < 42% in patients and 1-hour plasma glucose levels during rate of macrosomia, cesarean section, with GDM resulted in a decreased inci- a glucose challenge test compared to the fetal metabolic complications, shoulder dence of LGA infants, a decrease in diet-alone group.43 Another trial, in dystocia, neonatal intensive care unit cesarean deliveries for macrosomia and which women with GDM were random- days, and respiratory complications in cephalopelvic disproportion, and a ized to a partially home-based exercise the intensive management group. Anoth- decreased need for insulin therapy com- program, did not find any reduction in er important consideration of this study pared to patients on a diet with a higher blood glucose level, although the women is that GDM was defined as only one or carbohydrate content (45–50%).38 Con- did have an improvement in cardiovascu- more abnormal OGTT values, rather suming carbohydrates with a low lar fitness.40 Based on the potential bene- than the current standard of two or more glycemic index also results in lower fits of exercise in women with GDM, the abnormal glucose levels.46 Other studies postprandial glucose levels, especially ADA recommends starting or continuing have also shown improvement in rates of late in gestation.39 a program of moderate exercise in macrosomia and other maternal and fetal women without medical or obstetrical complications by treating women who Exercise contraindications.2 do not meet the criteria for GDM but The role of exercise in women with who have evidence of impaired carbohy- GDM has been controversial in the past Insulin drate tolerance as determined by an because maternal exercise on a bicycle Insulin therapy is the most commonly abnormal screening 50-g glucose chal- ergometer has been associated with fetal used treatment when MNT fails to lenge test and/or one or more abnormal bradycardia. Subsequent small stud- maintain blood glucose levels at the results on OGTT.47–49 ies40,41 have shown small transient desired ranges or when there is evidence Because there are no data demon- increases in fetal heart rate after mater- of excessive fetal growth. Small studies strating an optimal insulin regimen, the nal exercise. There were no fetal com- have demonstrated a decrease in macro- type and dose of insulin must be tailored plications in either study. somia44,45 as well as related morbidities to meet each patient’s requirements.3,27 Durak et al.42 found that uterine including operative deliveries and birth Human insulin is currently recommend- activity, defined as contractions with an trauma in women with GDM who were ed by the ADA.2,50 However, one study external tocometer deflection of > 15 treated with insulin.45 of 42 women with GDM diagnosed at mmHg above baseline for > 30 seconds, A large, prospective, population- 14–32 weeks of gestation found that varied in response to different types of based study of almost 2,500 women with insulin lispro was as effective as regular aerobic exercise, even at comparable lev- GDM compared the effect of intensive insulin in controlling glucose levels with els of exertion. The bicycle ergometer, versus conventional management of fewer episodes of hypoglycemia. Anti- treadmill, and rowing ergometer led to GDM. The women randomized to the insulin antibody levels were similar in uterine activity in 50, 40, and 10% of intensive management group were given the two groups. Additionally, the results exercise sessions, respectively. The memory reflectance meters and instruct- of umbilical cord blood in four patients recumbent bicycle and upper body ed to monitor their blood glucose seven who received continuous intravenous ergometer did not lead to any increase in times per day (fasting, preprandial, 2- insulin lispro and dextrose infusions uterine activity. Therefore, the authors hour postprandial, and bedtime). The intrapartum to assess placental insulin concluded that the recumbent bicycle women in the conventional management transfer did not detect any insulin and upper body ergometer were the group were instructed to monitor four lispro.51 Although insulin lispro appears safest forms of aerobic exercise for preg- times per day (fasting and 2-hour post- to be safe in pregnancy if started after 14 nant women. In addition, they recom- prandial) in addition to weekly fasting weeks of gestation, it is considered to be mended teaching women to palpate their and 2-hour postprandial glucose meas- in Pregnancy Category B by the Food uterus during exercise to detect subclini- urements during clinic visits. Both and Drug Administration (FDA), and the cal contractions and to discontinue the groups were treated with diet and insulin official recommendation of the ADA is exercise if contractions occur. as needed to reach the following goals: to use human insulin until further studies A potential benefit of exercise in overall mean blood glucose 90–100 verify the safety of insulin lispro.2 women with GDM is improved mg/dl, fasting blood glucose 60–90 The short-term efficacy of insulin glycemic control. One small trial ran- mg/dl, and postprandial blood glucose aspart was evaluated in a small study of domized women with GDM to diet and < 120 mg/dl. Overall, 66% of the women 15 women with GDM during standard- exercise with an arm ergometer versus in the intensive management group were ized meal tests. Although this study diet alone for 6 weeks. Researchers treated with insulin versus 36% of found that insulin aspart was effective in

20 Volume 23, Number 1, 2005 • CLINICAL DIABETES FEATURE ARTICLE

decreasing postprandial glucose concen- safety.2 Another potential concern with circumference > 70th percentile at 30 tration, further studies need to be done to the use of glyburide in GDM is possible weeks of gestation was associated with ensure the safety of this medication in impairment of myocardial ischemic pre- an increased risk of macrosomia.63 A pregnant women.52 Insulin aspart is con- conditioning.60 subsequent trial randomized 199 women sidered to be in Pregnancy Category C Metformin has also been used to to management based on maternal by the FDA. treat pregnant women with GDM. A ret- glycemia alone or glycemia plus ultra- The use of insulin glargine in humans rospective cohort study found an sound. The glucose thresholds for the has only been reported in case reports. increased prevalence of preeclampsia initiation of insulin differed in the two There have been no clinical trials evaluat- and perinatal mortality in women treated groups. The thresholds in the glycemia- ing the use of insulin glargine in pregnan- with metformin. However, the women in alone group were fasting glucose repeat- cy. It is currently considered to be in the metformin group were more obese edly > 90 mg/dl or 2-hour postprandial Pregnancy Category C by the FDA.53 and older, and their treatment was begun glucose > 120 mg/dl. The thresholds in later in gestation.57 Recent studies the glycemia-plus-ultrasound group were Oral Agents involving women with polycystic ovary fasting glucose > 120 mg/dl, 2-hour Currently, oral hypoglycemic agents are syndrome or women with type 2 dia- postprandial glucose > 200 mg/dl, or not recommended by the ADA or betes who continue metformin in preg- fetal abdominal circumference > 75th ACOG.2,27 The older sulfonylureas nancy have found no adverse pregnancy percentile. Ultrasound examinations chlorpropamide and tolbutamide could outcomes.61,62 were performed at entry and every 4 cross the , stimulate the fetal Although previous studies have been weeks starting at 20 weeks of gestation. pancreas, and cause fetal hyperinsuline- small, there is an ongoing prospective, Researchers found that neonatal out- mia.22,54 However, the transfer of gly- randomized controlled trial in New comes were equivalent and proposed that buride, a second-generation sulfony- Zealand and Australia comparing met- including fetal growth in the assessment lurea, across the human placenta was formin with insulin in women with of women with GDM may decrease glu- insignificant in experimental models.54,55 GDM. This study will help to answer cose testing in low-risk pregnancies. This finding led to a clinical trial of questions about the safety of metformin Therefore, antepartum fetal assessment 404 women with GDM randomized to during pregnancy.62 Metformin is listed with ultrasound may play a role in the either glyburide or insulin therapy at as Pregnancy Category B by the FDA. future management of patients with 11–33 weeks of gestation. There were no GDM.64 significant differences in glycemic con- ANTEPARTUM FETAL trol or adverse fetal outcomes. In addi- ASSESSMENT PERIPARTUM CONSIDERATIONS tion, glyburide was not detected in the ACOG recommends antepartum fetal When glycemic control is acceptable cord serum of any infants in the gly- assessment in women whose blood glu- and there are no other known complica- buride group.56 cose is poorly controlled, who require tions, routine delivery before 40 weeks Smaller studies have also supported insulin therapy, who have a history of an of gestation is not recommended.27 One the safety of glyburide use in pregnan- adverse obstetrical event, or who have a randomized trial of women with insulin- cy.57–59 In one of these trials, women with history of a hypertensive disorder. treated diabetes (93% of whom had GDM who were treated with glyburide Providers can determine which type of GDM) found that although induction of had fewer asymptomatic hypoglycemic antepartum test to use (biophysical pro- labor at 38 weeks of gestation resulted episodes compared to women with file, , or contraction stress in a smaller proportion of infants who GDM treated with insulin, although the test). were large for gestational age, there was clinical significance of these hypo- The role of antepartum testing in no difference in the rates of cesarean glycemic episodes is unknown.59 women with well-controlled GDM is delivery or shoulder dystocia.65 If a Thus, although glyburide appears to less clear.27 The recommendations of the delivery is indicated before 39 weeks, be safe in pregnancy based on the above Fourth International Workshop-Confer- pulmonary maturity should be assessed studies, it is important to recognize that ence on Gestational Diabetes Mellitus by before induction if these studies in aggregate are small and are to consider nonstress testing starting possible.27 not adequately powered to detect clini- at 32 weeks of gestation in patients on The rate of cesarean deliveries is cally important, relatively rare outcomes insulin and at or near term in patients much higher in women with GDM com- in pregnancy. Furthermore, glyburide is managed by diet alone.3 pared to women without GDM. The considered to be in Pregnancy Category Recent trials have assessed the use- increase in rate is higher than would be C by the FDA, and therefore it is not cur- fulness of fetal ultrasounds to help guide expected based solely on the associated rently recommended by the ADA or the management of patients with GDM. obstetric complications. Therefore, part ACOG until larger studies confirm its One study found that a fetal abdominal of this increase is likely influenced by

CLINICAL DIABETES • Volume 23, Number 1, 2005 21 FEATURE ARTICLE

physician knowledge of a history of fall between these thresholds. All Gestational Diabetes Mellitus. Diabetes Care 21 GDM.13 ACOG recommends counseling patients with a history of GDM should (Suppl. 2):B161–B167, 1998 women about the possibility of cesarean be educated about MNT, exercise, 4Brody SC, Harris R, Lohr K: Screening for gestational diabetes: a summary of the evidence section without labor when the estimated maintenance of normal body weight, the for the U.S. Preventive Services Task Force. fetal weight is > 4,500 g. If the estimated need for , and symp- Obstet Gynecol 101:380–392, 2003 fetal weight is 4,000–4,500 g, additional toms suggestive of hyperglycemia. 5Carpenter MW, Coustan DR: Criteria for screening tests for gestational diabetes. Am J risk factors for shoulder dystocia, such Obstet Gynecol 144:768–773, 1982 as clinical pelvimetry, progression of CONCLUSION 6Schmidt MI, Duncan BD, Reichelt AJ, labor, and patient’s past delivery history, GDM is a common medical problem Branchtein L, Matos MC, Forti A, Spichler ER, should be considered.27 that results from an increased severity of Pousada J, Teixeira MM, Yamashita T, for the Brazilian Gestational Diabetes Study Group: insulin resistance as well as an impair- Gestational diabetes mellitus diagnosed with a 2- POSTPARTUM CONSIDERATIONS ment of the compensatory increase in h 75-g oral glucose tolerance test and adverse pregnancy outcomes. Diabetes Care Women with GDM have an increased insulin secretion. Pregnancy, in essence, 24:1151–1155, 2001 risk of developing diabetes, most com- serves as a metabolic stress test and 7Cianni GD, Miccoli R, Volpe L, Lencioni C, monly type 2 diabetes, after pregnancy. uncovers underlying insulin resistance Del Prato S: Intermediate metabolism in normal pregnancy and in gestational diabetes. Diabetes Although follow-up studies on the and -cell dysfunction. GDM is associ- Metab Res Rev 19:259–270, 2003 cohort of patients used to derive the ated with a variety of maternal and fetal 8Xiang AH, Peters RK, Trigo E, Kjos SL, Lee O’Sullivan and Mahan criteria for GDM complications, most notably macroso- WP, Buchanan TA: Multiple metabolic defects found diabetes in 50% of women who mia. during late pregnancy in women at high risk for type 2 diabetes. Diabetes 48:848–854, 1999 had previously had GDM,27 the reported Controversy surrounds the ideal 9Maurico D, Balsells M, Morales J, Corcoy R, prevalence varies. A recent systematic approach for detecting GDM, and the Puig-Domingo M, de Leiva A: Islet cell autoim- literature review of 28 studies found that approaches recommended for screening munity in women with gestational diabetes and risk of progression to insulin-dependent diabetes the cumulative incidence of type 2 dia- and diagnosis are largely based on expert mellitus. Diabetes Metab Rev 12:275–285, 1996 betes ranged from 2.6 to > 70% in stud- opinion. Controlling maternal glycemia 10Kjos AL, Buchanan TA: Gestational dia- ies with postpartum follow-up ranging with MNT, close monitoring of blood betes mellitus. N Engl J Med 341:1749–1756, from 6 weeks to 28 years.66 A meta- glucose levels, and treatment with 1999 analysis calculated the relative risk for insulin if blood glucose levels are not at 11Casey BM, Lucas MJ, McIntire DD, Leveno KJ: Pregnancy outcomes in women with gesta- developing diabetes after GDM to be 6.0 goal has been shown to decrease fetal tional diabetes compared with the general obstet- (95% CI 4.1–8.8).67 and maternal morbidities. In addition, ric population. Obstet Gynecol 90:869–873, 1997 Factors associated with an increased certain types of exercise appear to have 12Dang K, Homko C, Reece AE: Factors asso- risk of progression to diabetes within 5 potential benefits in women without any ciated with fetal macrosomia in offspring of ges- tational diabetic women. J Matern Fetal Med years of the diagnosis of GDM include contraindications. 9:114–117, 2000 gestational age at diagnosis, the level of Other treatment modalities, such as 13Naylor CD, Phil D, Sermer M, Chen E, glycemia at diagnosis and at the first oral agents, need further study to validate Sykora K, for The Toronto Tri-Hospital Gesta- tional Diabetes Investigators: Cesarean delivery postpartum assessment, impairment of their safety and efficacy. Additionally, in relation to birth weight and gestational glucose -cell function, obesity, and further more research on the use of antepartum tolerance: pathophysiology or practice style? JAMA pregnancy.3 Ethnicity may also be a risk fetal assessment to help guide treatment 275:1165–1170, 1999 factor for progression to diabetes. How- in women with GDM is needed. 14O’Sullivan JB, Charles D, Mahan CM, Dandrow RV: Gestational diabetes and perinatal ever, further studies are needed to clarify Finally, postpartum management of mortality rate. Am J Obstet Gynecol 116:901– this issue.66 women with GDM is critical because of 904, 1973 Maternal glycemic status should be their markedly increased risk of type 2 15Beischer NA, Wein P, Sheedy MT, Steffen B: Identification and treatment of women with reclassified 6 weeks or more after preg- diabetes in the future. hyperglycaemia diagnosed during pregnancy can nancy ends and every 3 years thereafter significantly reduce perinatal mortality rates. Aust as either diabetes mellitus, impaired REFERENCES NZ J Obstet Gynecol 36:239–247, 1996 fasting glucose, impaired glucose toler- 16Wood SL, Sauve RS, Ross S, Brant R, Love 1Expert Committee on the Diagnosis and E: Prediabetes and perinatal mortality. Diabetes ance, or normoglycemia.1 Normal val- Classification of Diabetes Mellitus: Report of the Care 23:1752–1754, 2000 expert committee on the diagnosis and classifica- ues for a 2-hour OGTT are fasting 17 tion of diabetes mellitus. Diabetes Care 26 (Sup- Girz BA, Divon MY, Merkatz IR: Sudden < 100 mg/dl and 2-hour post–75-g glu- pl. 1):S5–S20, 2003 fetal death in women with well-controlled, inten- sively monitored gestational diabetes. J Perinatol cose load < 140 mg/dl. Glucose values 2American Diabetes Association: Gestational 12:229–233, 1992 that meet the criteria for diabetes melli- diabetes mellitus ( Statement). Diabetes 18 Care 27 (Suppl. 1):S88–S90, 2004 Kjos SL, Leung A, Henry OA, Victor MR, tus are > 126 mg/dl or 2-hour post–glu- Paul RH, Medearis AL: Antepartum surveillance cose load > 200 mg/dl. Impaired fasting 3Metzger BE, Coustan DM, Organizing Com- in diabetic pregnancies: predictors of fetal dis- mittee: Summary and recommendations of the tress in labor. Am J Obstet Gynecol 175:1532– glucose and impaired glucose tolerance Fourth International Workshop-Conference on 1539, 1995

22 Volume 23, Number 1, 2005 • CLINICAL DIABETES FEATURE ARTICLE

19Sheffield JS, Butler-Koster EL, Casey BM, ples and recommendations for the treatment and test) decreases the prevalence of macrosomia. Am McIntire DD, Leveno KJ: Maternal diabetes mel- prevention of diabetes and related complications J Perinat 16:269–275, 1999 litus and infant malformations. Obstet Gynecol (Technical Review). Diabetes Care 25:148–198, 48 100:925–930, 2002 2002 Sermer M, Naylor CD, Phil D, Gare DJ, Kenshole AB, Ritchie JWK, Farine D, Cohen 20Schaefer UM, Songster G, Xiang A, 33Jovanovic-Peterson L, Peterson CM: Nutri- HR, McArthur K, Holzapfel S, Biringer A, Chen Berkowitz K, Buchanan TA, Kjos SL: Congenital tional management of the obese gestational dia- E, for the Toronto Tri-Hospital Gestational Dia- malformations in offspring of women with hyper- betic pregnant woman. J Am Coll Nutr betes Investigators: Impact of increasing carbohy- glycemia first detected during pregnancy. Am J 11:246–250, 1992 drate intolerance on maternal-fetal outcomes in

Obstet Gynecol 177:1165–1171, 1997 34 3,637 women without gestational diabetes. Am J Knopp RH, Magee MS, Raisys V, Benedetti Obstet Gynecol 173:146–156, 1995 21Seely EW, Solomon CG: Insulin resistance T: Metabolic effects of hypocaloric diets in man- and its potential role in pregnancy-induced hyper- agement of gestational diabetes. Diabetes 40 49Langer O, Anyaegbunam A, Brustman L, tension. J Clin Endocrinol Metab 88:2393–2398, (Suppl. 2):165–171, 1991 Divon M: Management of women with one 2003 35 abnormal oral glucose tolerance test value Rizzo T, Metzger BE, Burns WJ, Burns K: Am J 22Naylor CD, Phil D, Sermer M, Chen E, Correlations between antepartum maternal reduces adverse outcome in pregnancy. Farine D: Selective screening for gestational dia- metabolism and intelligence of offspring. N Engl Obstet Gynecol 161:593–599, 1989 betes mellitus. N Engl J Med 337:1591–1596, J Med 323:911–916, 1991 50Jovanovic-Peterson L, Kitzmiller JL, Peter- 1997 36Rizzo TA, Dooley SL, Metzger BE, Cho son CM: Randomized trial of human versus ani- 23Mikola M, Hiilesmaa V, Halttunen M, Suho- NH, Ogata ES, Silverman BL: Prenatal and peri- mal species insulin in diabetic pregnant women: nen L, Tiitinen A: Obstetric outcome in women natal influences on long-term psychomotor devel- improved glycemic control, not fewer antibodies with polycystic ovarian syndrome. Hum Reprod opment in offspring of diabetic mothers. Am J to insulin, influences birth weight. Am J Obstet 16:226–229, 2001 Obstet Gynecol 173:1753–1758, 1995 Gynecol 167:1325–1330, 1992 24Bjercke S, Dale PO, Tanbo T, Storeng R, 37Peterson CM, Jovanovic-Peterson L: Per- 51Jovanovic L, Ilic S, Pettitt DJ, Hugo K, Ertzeid G, Abyholm T: Impact of insulin resist- centage of carbohydrate and glycemic response to Gutierrez M, Bowsher RR, Bastyr EJ: Metabolic ance on pregnancy complications and outcome in breakfast, lunch, and dinner in women with ges- and immunologic effects of insulin lispro in ges- women with polycystic ovary syndrome. Gynecol tational diabetes. Diabetes 40 (Suppl. tational diabetes. Diabetes Care 22:1422–1427, Obstet Invest 54:94–98, 2002 2):172–174, 1991 1999 25Vollenhoven B, Clark S, Kovacs G, Burger 38Major CA, Henry MJ, De Veciana M, Mor- 52Pettitt DJ, Ospina P, Kolaczynski JW, H, Healy D: Prevalence of gestational diabetes gan MA: The effects of carbohydrate restriction Jovanovic L: Comparison of an insulin analog, mellitus in polycystic ovarian syndrome (PCOS) in patients with diet-controlled gestational dia- insulin aspart, and regular human insulin with no patients pregnant after ovulation induction with betes. Obstet Gynecol 91:600–604, 1998 insulin in gestational diabetes mellitus. Diabetes gonadotrophins. Aust NZ J Obstet Gynecol 39 Care 26:183–186, 2003 40:54–59, 2000 Clapp JF: Effect of dietary carbohydrate on the glucose and insulin response to mixed caloric 53 26 Devlin JT, Hothersall L, Wilkis JL: Use of Haakova L, Cibula D, Rezabek K, Hill M, intake and exercise in both nonpregnant and preg- insulin glargine during pregnancy in a type 1 dia- Fanta M, Zivny J: Pregnancy outcome in women nant women. Diabetes Care 21 (Suppl. betic woman. Diabetes Care 25:1095–1096, 2002 with PCOS and in controls matched by age and 2):B107–B112, 1998 weight. Hum Reprod 18:1438–1441, 2003 54 40Avery MD, Leon AS, Kopher RA: Effects of Elliot BD, Schenker S, Langer O, Johnson RF, Prihoda T: Comparative placental transport of 27 a partially home-based exercise program for American College of and Gyne- oral hypoglycemic agents in humans: a model of cologists Committee on Practice Bulletins— women with gestational diabetes. Obstet Gynecol 89:10–15, 1997 human placental drug transfer. Am J Obstet Obstetrics: Gestational diabetes. Number 30, Gynecol 171:653–660, 1994 September 2001. Obstet Gynecol 98:525–538, 41 Collings C, Curet LB: Fetal heart rate 55 2001 response to maternal exercise. Am J Obstet Elliot BD, Langer O, Schenker S, Johnson Gynecol 151:498–501, 1985 RF: Insignificant transfer of glyburide occurs 28Jovanovic-Peterson L, Bevier W, Peterson across the human placenta. Am J Obstet Gynecol CM, the Santa Barbara County Health Care Ser- 42Durak EP, Jovanovic-Peterson L, Peterson 165:807–812, 1991 vices Program: Birth weight change concomitant CM: Comparative evaluation of uterine response 56 with screening for and treatment of glucose-intol- to exercise on five aerobic machines. Am J Obstet Langer O, Conway DL, Berkus MD, erance of pregnancy: a potential cost-effective Gynecol 162:754–756, 1990 Xenakis E, Gonzales O: A comparison of gly- intervention? Am J Perinat 14:221–227, 1997 buride and insulin in women with gestational dia- 43Jovanovic-Peterson L, Durak EP, Peterson betes mellitus. N Engl J Med 343:1134–1138, 29Langer O, Levy J, Brustman L, Anyaeg- CM: Randomized trial of diet versus diet plus 2000 bunam A, Merkatz R, Divon M: Glycemic control cardiovascular conditioning on glucose levels in 57 in gestational diabetes mellitus: how tight is tight gestational diabetes. Am J Obstet Gynecol Hellmuth E, Damm P, Molsted-Pedersen: enough: small for gestational age versus large for 161:415–419, 1989 Oral hypoglycaemic agents in 118 diabetic preg- gestational age? Am J Obstet Gynecol nancies. Diabet Med 17:507–511, 2000 44 161:646–653, 1989 Thompson DJ, Porter KB, Gunnells DJ, 58 Wagner PC, Spinnato JA: Prophylactic insulin in Kremer CJ, Duff P: Glyburide for the treat- 30Jovanovic-Peterson L, Peterson CM, Reed the management of gestational diabetes. Obstet ment of gestational diabetes. Am J Obstet GF, Metzger BE, Mills JL, Knopp RH, Aarons Gynecol 75:960–964, 1990 Gynecol 190:1438–1439, 2004 JH: Maternal postprandial glucose levels and 59 infant birth weight: the Diabetes in Early Preg- 45Coustan DR, Imarah J: Prophylactic insulin Yogev Y, Ben-Haroush A, Chen R, Rosenn nancy Study. Am J Obstet Gynecol 164:103–111, treatment of gestational diabetes reduces the inci- B, Hod M, Langer O: Undiagnosed asymptomatic 1991 dence of macrosomia, operative delivery and hypoglycemia: diet, insulin, and glyburide for birth trauma. Am J Obstet Gynecol 150:836–842, gestational diabetic pregnancy. Obstet Gynecol 31De Veciana M, Major CA, Morgan MA, 1984 104:88–93, 2004 Asrat T, Toohey JS, Lien JM, Evans AT: Post- 60 prandial versus preprandial blood glucose moni- 46Langer O, Rodriguez DA, Xenakis E, Riddle MC: Sulfonylureas differ in effects toring in women with gestational diabetes melli- McFarland MB, Berkus MD, Arredondo F: Inten- on ischemic preconditioning: is it time to retire tus requiring insulin therapy. N Engl J Med sified versus conventional management of gesta- glyburide? J Clin Endocrinol Metab 88:528–530, 333:1237–1241, 1995 tional diabetes. Am J Obstet Gynecol 2003 170:1036–1047, 1994 61 32Franz MJ, Bantle JP, Beebe CA, Brunzell Glueck CJ, Wang P, Goldenberg N, Sieve- JD, Chiasson JL, Garg A, Holzmeister LA, Hoog- 47Bevier WC, Fischer R, Jovanovic L: Treat- Smith L: Pregnancy outcomes among women werf B, Mayer-Davis E, Mooradian AD, Purnell ment of women with an abnormal glucose chal- with polycystic ovary syndrome treated with met- JQ, Wheeler M: Evidence-based nutrition princi- lenge test (but a normal oral glucose tolerance formin. Hum Reprod 17:2858-2864, 2002

CLINICAL DIABETES • Volume 23, Number 1, 2005 23 FEATURE ARTICLE

62Simmons D, Walters B, Rowan JA, McIn- evaluating a predominately fetal growth–based nificance of gestational diabetes. Diabetes Care tyre HD: Metformin therapy and diabetes in preg- strategy to guide management of gestational dia- 26:2005–2009, 2003 nancy. Med J Aust 180:462–464, 2004 betes in Caucasian women. Diabetes Care 27:297–302, 2004 63Kjos SL, Schaefer-Graf U, Sardesi S, Peters RK, Buley A, Xiang AH, Byrne JD, Sutherland 65Kjos SL, Henry OA, Montoro M, Buchanan Tracy L. Setji, MD, is an endocrinology C, Montoro MN, Buchanan TA: A randomized TA, Mestman JH: Insulin-requiring diabetes in fellow; Ann J. Brown, MD, is an assis- controlled trial utilizing glycemic plus fetal ultra- pregnancy: a randomized trial of active induction sound parameters versus glycemic parameters to of labor and expectant management. Am J Obstet tant professor of medicine; and Mark N. determine insulin therapy in gestational diabaetes Gynecol 169:611–615, 1993 Feinglos, MD, CM, is a professor of with fasting hyperglycemia. Diabetes Care 66 medicine and chief of endocrinology in 24:1904–1910, 2001 Kim C, Newton KM, Knopp RH: Gestation- al diabetes and the incidence of type 2 diabetes. the Department of Medicine, Division of 64Schaefer-Graf UM, Kjos SL, Fauzan OH, Diabetes Care 25:1862–1868, 2002 Endocrinology, at Duke University Med- Buhling KJ, Siebert G, Buhrer C, Ladendorf B, 67 Dudenhausen JW, Vetter K: A randomized trial Cheung NW, Byth K: Population health sig- ical Center in Durham, N.C.

24 Volume 23, Number 1, 2005 • CLINICAL DIABETES