Abstract
The American Diabetes Association sponsored the 5th International Workshop–Conference on Gestational Diabetes, held on November 11–13, 2005, in Chicago (IL, USA). The conference was chaired by Boyd Metzger of Northwestern University Feinberg School of Medicine (IL, USA). The meeting provided an opportunity for clinicians and researchers from all over the world to review new evidence-based clinical and basic information concerning gestational diabetes mellitus (GDM). An overview of many of the formal lectures of the conference is presented here.
Key presentations
Epidemiology
Gestational diabetes is defined as ‘glucose intolerance with onset or first recognition during pregnancy’. This diagnosis sometimes captures patients with previously undiagnosed diabetes (usually Type 2 diabetes mellitus [DM]), particularly when individuals have had limited medical care prior to pregnancy. Population-wide data on the prevalence of GDM are very limited, as the condition is not reportable and screening and diagnostic approaches are not standardized or applied in a uniform manner. Assiamira Ferrara, of the University of Washington (WA, USA), and the Kaiser Foundation Research Institute, Oakland (CA, USA), presented results from two recent reports, derived from data in large healthcare plans, that provided convincing evidence that the prevalence of GDM has increased by 40–50% in the USA in the last decade. This increase in GDM temporally coincides with what is widely viewed as an epidemic of obesity, commonly considered to be a national emergency. All racial/ethnic groups are similarly affected. Between 1991 and 2001, the prevalence of GDM also increased by 50% among the youngest cohort of women. GDM affects 5–8% of all pregnancies in the USA, where the risk of developing Type 2 diabetes in the 5–10 years following a pregnancy is 50% or more, with a lifetime risk of approximately 80%.
Ultimately, these perinatal and long-term consequences of GDM will have an impact on all who practice medicine; thus, it is imperative that GDM be addressed in the fullest context and not as merely an obstetrical dilemma. GDM potentially increases the morbidity of the sentinel pregnancy during which it is diagnosed, and may alter the lifelong health of the mother as well as her offspring. Babies born to mothers with GDM are themselves at an increased risk of childhood obesity and early-onset Type 2 diabetes.
A number of studies currently underway have sought to elucidate the pathophysiology of this condition in the hope of decreasing its prevalence and associated morbidity. Boyd Metzger presented a progress report on the Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study, which is scheduled to be completed in 2007. The primary objective of the HAPO study is to clarify unanswered questions on the association of maternal hyperglycemia, which is less severe than overt DM, with the risk of adverse pregnancy outcome. The development of outcome-based criteria for the diagnosis and classification of GDM is expected to follow, perhaps drafted at the 6th International Workshop–Conference.
Pathophysiology
The etiology of GDM is multifactorial, as the diagnosis encompasses all cases of ‘hyperglycemia with onset or first recognition during pregnancy’ and identifies subjects with more than one disease entity. In the majority of cases, GDM can be viewed as the forerunner of Type 2 DM. David Dunger, University of Cambridge (UK), pointed out that the detection process also captures individuals with rare disorders (e.g., monogenic forms of diabetes characterized by genetic alterations in β-cell function, or mitochondrial transmitted diabetes), as well as some individuals with early, evolving autoimmune diabetes (Type 1 DM).
Thomas Buchanan, of the University of Southern California Keck School of Medicine (CA, USA), reviewed data demonstrating that women who develop GDM tend to have greater than normal insulin resistance and defects in β-cell function that can be demonstrated outside of pregnancy. During pregnancy, there is a further increase in insulin resistance in healthy pregnancies and in pregnancies complicated by GDM. Subjects with normal β-cells increase insulin secretion to compensate and retain normal circulating glucose concentrations. Those who develop GDM do increase insulin secretion as insulin resistance increases, but since their functioning β-cell mass is reduced, hyperglycemia ensues. Jed Friedman, University of Colorado Health Science Center (CO, USA), summarized his research and that of others demonstrating some of the intracellular defects that mediate insulin resistance in normal pregnancy and GDM.
Alberto de Leiva, Universitat Automnoma of Barcelona (Spain), discussed potential autoimmune contributions in a subset of women found to have GDM. In this scenario, a patient loses β-cells due to an autoimmune insult and, consequently, insulin secretion is not sufficient to compensate for the insulin resistance of pregnancy. In the postnatal period, patients who have been demonstrated to have islet autoantibodies follow one of three paths: restoration of normal glucose tolerance, rapid progression to Type 1 DM or gradual deterioration of insulin secretion as a result of a more slowly progressing autoimmune process, eventually culminating in Type 1 DM. In Leiva's experience, approximately 10% of Caucasian women with GDM have demonstrable autoimmune alterations.
Sylvie Hauguel-de Mouzon, MetroHealth Medical Center (OH, USA), and Gernot Desoye, Medical University of Graz (Austria), discussed potential roles that the placenta may play in the overall morbidity of GDM. According to these researchers, the placenta responds to hormonal and metabolic cues produced by both the fetus and the mother. The timing of the diabetic insult and its severity dictate the type and degree of the response in the placenta. Altered maternal nutrients, metabolic factors or signaling molecules from the mother may have a direct impact on placental metabolism and function. However, signals from the fetus that result from the alterations in its metabolic environment may also feedback to the placenta and influence its function.
Offspring of mothers with gestational diabetes
Patrick Catalano, MetroHealth Medical Center (OH, USA), discussed the determinants of birth weight in general and in GDM specifically. In the general population, maternal prepregnancy body mass index (BMI) is highly significantly correlated with the birth-weight of offspring. He presented data that demonstrated a gradual and consistent increase in the birthweight of babies born over the last 30 years in the MetroHealth Medical Center. The increase is seen across the spectrum from the lowest to highest birthweight categories. Thus, what would now be large for gestational age (LGA) according to percentile birthweight distribution is higher than in the past. This rise in birth weight parallels an increase in the average population BMI. Catalano also discussed practical concerns involving this upward drift in infant birthweight, as it relates to GDM. Infants born to mothers with GDM, regardless of their actual weight, have increased body fat and decreased lean body mass when compared with other infants. The fat distribution, not the actual weight, is the source of maternal and neonatal morbidity such as birth trauma. For this reason, calculations that convey body composition, such as the ponderal index, represent a better indication of the metabolic effects of diabetes and GDM than absolute fetal weight.
Erica Gunderson, Kaiser Foundation Research Institute (CA, USA), discussed the pros and cons of breastfeeding after a pregnancy complicated by GDM. In nondiabetic patients, breastfeeding may promote postpartum weight loss and is associated with a decreased incidence of Type 2 DM in offspring, as well as lower prevalences of childhood and adolescent obesity. While clear neonatal and maternal benefits of breastfeeding have been well established in nondiabetic populations, studies in patients with GDM are lacking. Studies in Type 1 DM patients demonstrated a reduction in rates of Type 1 diabetes in genetically susceptible infants who were breastfed when compared with those who were formula fed. Conversely, a study from Berlin (Germany) has been published that found more obesity and impaired glucose tolerance in children of Type 1 DM mothers who received breast milk from their mothers than in those fed breast milk from nondiabetic donors. There is a need for studies addressing the affect of breastfeeding in patients with GDM, particularly as it relates to the development of postpartum obesity, impaired glucose tolerance and Type 2 DM. It is unclear whether there are metabolic effects, such as less retention of weight gained during pregnancy, that may occur as a result of breastfeeding in this population. Given the benefits of breastfeeding in the general population, it is considered appropriate to encourage breastfeeding by mothers after GDM unless studies show definitive evidence of significant adverse consequences.
Management of gestational diabetes
Metabolic profiles in normal pregnancy
The therapeutic objective in GDM is the restoration of normal metabolism. However, studies of diurnal metabolic profiles in normal pregnancy are few in number and have typically involved a small number of subjects studied in a controlled setting, such as a clinical research center. Moshe Hod, Rabin Medical Center, Tel Aviv (Israel), presented data from continuous glucose monitoring over a 72 h interval in 57 pregnant women with normal glucose metabolism. In third-trimester subjects with normal glucose tolerance, glucose levels increased from fasting (75 ± 12 mg/dl standard deviations [SD]) or premeal (78 ± 11 mg/dl) to peak concentrations (110 ± 16 mg/dl) approximately 70 min after mixed-nutrient meals. In diabetic patients (diet-treated GDM patients, insulin-treated GDM or Type 1 DM), these peaks were later and higher. In diet-treated GDM, values were: fasting (84 ± 18 mg/dl) and postprandial peak (131 ± 29 mg/dl), with the peak occurring 82 ± 31 min after initiating meals. Hod pointed out that GDM is characterized by multiple metabolic defects in addition to hyperglycemia; for example, alterations in other fuel sources such as lipids and amino acids. Investigators are well aware of these features. However, in usual clinical practice, only glucose levels are monitored since it is simple to do, and similar techniques are not available to monitor other metabolic parameters.
Obstetrics management
Decisions concerning timing and route of delivery for pregnancies complicated by GDM are often based on assessments of fetal weight. Ultrasound estimates are less accurate at the upper and lower extremes of fetal size; therefore, the sensitivity and specificity of identifying large fetuses is not optimal. Deborah Conway, University of Texas Health Science Center (TX, USA), presented results from a number of small studies evaluating the use of magnetic resonance imaging (MRI) and 3D ultrasound technology to obtain volumetric calculations of fetal size. Fetal weight obtained via these modalities was consistently within 5% of actual weight. Unfortunately, these studies excluded patients with abnormalities of fetal growth, including patients with suspected macrosomia. In order to fully assess the utility of MRI and 3D ultrasound in the management of obstetric patients, larger and more inclusive studies are needed.
There is a positive correlation between fetal size and risk of fetal birth trauma. Some clinicians advocate elective cesarean delivery in patients with suspected macrosomia in order to avoid these adverse outcomes. Maternal morbidity, including risk of hemorrhage and wound infection, are increased with surgical delivery when compared with vaginal delivery. Furthermore, patients who undergo multiple cesarean deliveries experience increased morbidity with each successive surgical delivery. The overall rate of severe brachial plexus injury is less than 1% of all pregnancies; therefore, the number of cesarean deliveries performed to prevent a single serious injury is quite high. There is insufficient evidence regarding the balance of maternal risk and fetal benefit to support a guideline for performing elective cesarean deliveries. Nonetheless, this approach may be warranted for fetuses of the highest estimated birth weight (i.e., more than 4500 g), especially if other ultrasound parameters, such as abdominal circumference, suggest excessive fetal fat deposition.
Fetal-based therapy
Siri Kjos, Harbor-UCLA Medical Center (CA, USA), and colleagues have carried out studies that have used estimates of fetal size and growth rate to guide the intensity of medical therapy. Kjos summarized results from multiple reports of studies that used fetal-based therapy. Treatment algorithms were based on fetal determinations, such as ultrasound measurements and amniotic fluid insulin levels. Patients showing neonatal effects of hyperglycemia, such as polyhydramnios or macrosomia, were treated with intensive insulin therapy and increased antenatal surveillance. Patients without these signs received more liberal surveillance and were managed with dietary as opposed to medical therapy. Neonatal outcome was similar in both groups, with no significant difference in birthweight or neonatal morbidity. The combination of suboptimal blood glucose control and third-trimester abdomen circumference measurements greater than the 75th percentile was associated with an increased risk of a LGA infant. Likewise, adequate blood glucose control and third-trimester abdomen circumference measurement less than the 75th percentile were associated with a decreased risk of a LGA infant. The collective findings of these studies suggest that these fetal markers are reliable indicators of fetal wellbeing that can be used to guide therapy for patients with GDM.
Medical nutritional therapy
Medical nutrition therapy (MNT) is the cornerstone of the management of GDM. As discussed by Diane Reader, International Diabetes Center (MN, USA), the overall goal of the management of GDM is a healthy pregnancy and baby. The objectives of MNT include control of carbohydrate intake and calories to facilitate adequate weight gain and normoglycemia in the absence of ketones. In order to support a healthy pregnancy, a woman typically requires a daily intake of approximately 30 kcal/kg body weight. A minimum of 175 g of carbohydrates should be included in the daily allowance. Patients should be encouraged to engage in 15–30 min of daily aerobic activity for optimal health. Only 30% of all pregnant women adhere to present guidelines for weight gain during pregnancy. A total of 50% of women gain more weight than is recommended and 20% fail to reach ideal goals.
Issues regarding MNT for GDM that offer important opportunities for research include how to safely and effectively manipulate energy intake, dietary composition and gestational weight gain. Similar to Type 2 DM, GDM is a state of insulin resistance and insulin deficiency. Weight control and exercise are modalities that improve insulin sensitivity and are effective in the treatment of Type 2 DM. The optimal application of these modalities in the management of GDM requires further study.
In the event that MNT and adequate physical activity are unable to achieve or sustain acceptable maternal glycemia, more intensified therapy may be necessary.
Pharmacological therapy
Insulin
Historically, when MNT and exercise do not control maternal glycemia optimally, the administration of insulin has been recommended for the treatment of GDM. Lois Jovanovic, Sansum Diabetes Research Institute (CA, USA), discussed the use of insulin therapy in pregnancy. Human insulin synthesized with the use of recombinant DNA is attended by minimal anti-insulin antibody levels and has therefore been the preferred form of therapy. Rapidly absorbed insulin analogs were developed to provide postinjection circulating insulin patterns that more closely mimic physiological postprandial profiles than those achieved with standard insulin therapy. At this time, three rapid-acting analogs are available; insulin lispro, insulin aspart and glulisine. To date, lispro and aspart have been found in clinical trials to be safe in GDM, with no increase in antibody titers and no increase in adverse pregnancy outcomes compared with subjects treated with regular human insulin, although relatively small numbers of subjects have been studied. Long-acting insulin analogs are now available for clinical use; however, none have been evaluated for use in GDM.
Oral medications
Glyburide
Interest in the use of oral medication for the treatment of GDM has increased widely since Langer and colleagues published a randomized clinical trial that compared the results of treatment of GDM with sulfonylurea, glyburide and insulin. Thomas Moore, UCSD School of Medicine (CA, USA), discussed the results of this original clinical trial and other published experiences with glyburide therapy. The study found no difference in average serum glucose concentration, or rates of macrosomia or LGA babies, and the transplacental transfer of glyburide was shown to be minimal. A total of 4% of the patients were deemed glyburide failures and were subsequently treated with insulin.
Since the completion of that study, four additional studies have been completed with similar results. A high proportion of glyburide-treated patients have had good glycemic control and fewer episodes of hypoglycemia. Treatment with glyburide during pregnancy was determined to be more cost effective than treatment with insulin.
Metformin
Interest has arisen regarding the use of metformin during pregnancy for the treatment of very insulin-resistant subjects with Type 2 DM or GDM. In contrast to glyburide, metformin readily crosses the placenta. A randomized clinical trial, Metformin in Gestational diabetes (MiG), is now underway in Australia and New Zealand. Janet Rowan, National Women's Hospital, Auckland (New Zealand), presented a report on the design and progress of this prospective, randomized, open-label, multicenter trial. The purpose of the study is to compare neonatal and maternal outcome variables in patients with GDM treated with insulin and metformin. Outcome variables to be assessed include episodes of hypoglycemia, rates of neonatal respiratory distress syndrome (RDS), hyperbiliru-binemia, birth trauma, Apgar scores, prematurity and neonatal intensive care unit admissions. It is anticipated that recruitment for the study will end in September 2006 with analysis of results to follow.
Postpartum care
Glucose metabolism
Since women with a history of GDM are at a higher risk of developing impaired glucose tolerance or Type 2 DM, the postpartum phase proves to be an exceptional opportunity to counsel patients regarding how to delay or prevent overt diabetes.
John Kitzmiller, UCSF (CA, USA), reported that patients who experienced GDM usually have normal glucose levels in the immediate postpartum phase. However, blood glucose levels should be assessed following delivery to identify the occasional patients with persisting diabetes (fasting levels ≥126 mg/dl or postprandial ≥200 mg/dl). Following delivery, women should be encouraged to continue MNT principles and increase physical activity to reduce weight gained during pregnancy. At 6–12 weeks postpartum, the recommended test to determine glucose abnormality should be the 2-h 75 g oral glucose tolerance test (OGTT) with fasting and 2-h values. The aim of the OGTT is to identify those with impaired fasting glucose (plasma glucose ≥ 100–125 mg/dl), impaired glucose tolerance (2-h plasma glucose 140–199 mg/dl) or diabetes (fasting plasma glucose ≥ 126 mg/dl; 2-h OGTT value ≥ 200 mg/dl). Those with normal glucose tolerance should be counseled regarding the high probability of developing GDM in subsequent pregnancies. All postpartum patients should be aware of the risks of Type 2 DM and cardiovascular disease. Multiple studies have confirmed the high risk of developing Type 2 DM within 5–10 years of having GDM. Since diabetes and potential cardiovascular risk factors may be present soon after pregnancy, there is a need for ongoing medical evaluation, disease prevention efforts and treatment for women with a previous history of GDM.
Robert Ratner, Medstar Research Institute (MD, USA), summarized the results of recent randomized, controlled trials that have examined potential approaches to the prevention of Type 2 diabetes following GDM. The subjects in these cohorts were at a very high risk of progression to DM. In one study, the Diabetes Prevention Program, lifestyle modification (e.g., weight loss, increasing daily activity, decreasing the amount of calories consumed on a daily basis) resulted in a greater than 50% reduction of the risk of developing DM. In the same population, metformin was equally effective in preventing DM in women who had previous GDM. In another study that was conducted in Los Angeles County (CA, USA) among Latino women with previous GDM, the TRoglitazone In Prevention Of Diabetes (TRIPOD) study, use of the medication troglitazone (subsequently withdrawn from use due to side effects) also resulted in a reduction in progression to DM after GDM of over 50%. A preliminary report from the follow-up study of the same cohort using another medication in the same class (pioglitazone), which does not have the same liver injury side effects, indicated that this medication is as effective as troglitazone.
Contraception
Contraception and pregnancy planning should be discussed at each annual healthcare visit during the reproductive years of women with a previous history of GDM. Siri Kjos, Harbor-UCLA Medical Center, Torrance (CA, USA), explained the many contraception options available to women. Prior to the initiation of any contraceptive method, patient education and maternal care are vital components to either prevent unwanted pregnancies or plan future pregnancies in the most optimal glycemic environment. Women with a previous history of GDM should be counseled to choose the most appropriate method of contraception to minimize the adverse affects of glucose intolerance and serum lipids. A long-term controlled study found that continued use of two combination oral contraceptives, one with monophasic norethindrone and the other with triphasic levonorgestrel, did not contribute to the development of diabetes compared with nonhormonal methods. However, progesterone-only oral contraceptive agents caused varying degrees of carbohydrate impairment and diabetes among breastfeeding women. Therefore, women with a prior history of GDM may not be candidates for the use of progesterone-only oral contraceptives while breastfeeding, but rather a nonhormonal or low-dose combined estrogen and progesterone method. Parenterally administered medroxyprogesterone should also be considered a second-line contraceptive choice.
Progesterone- and nonprogesterone-containing intrauterine devices (IUD) provide an effective option for postnatal contraception after GDM. A progesterone-containing IUD has not been found to exert systemic hormonal effects on carbohydrate metabolism. In appropriately selected subjects, an IUD is a very effective family planning option.
Preventive health management
Marshall Carpenter, Brown University, Providence (RI, USA), reviewed studies linking GDM with the risk of vascular disease later in life. GDM is associated with a higher risk of pre-eclampsia, which leads to vascular alterations that may later be manifested as coronary vascular disease. GDM is also associated with a metabolic syndrome that includes obesity, hyperlipidemia and hypertension. All of these disease processes contribute to coronary vascular disease. Improvements of insulin resistance through lifestyle modifications and, when necessary, pharmacological agents, have the potential to decrease the severity and perhaps the occurrence of other chronic conditions later in life.
Andrea Dunaif, Feinberg School of Medicine, Northwestern University, Chicago (IL, USA), discussed polycystic ovary syndrome (PCOS). Similarly to GDM, this syndrome exhibits obesity and insulin resistance, and is often associated with β-cell dysfunction and glucose intolerance. Among women with PCOS, the prevalence rate of GDM has been reported to be as high as 20% in the first trimester of pregnancy and 30% at 32 weeks of pregnancy. PCOS is the leading cause of anovulatory infertility. The drug of choice to stimulate ovarian production is metformin. At present, results are not available from randomized clinical trials to support the use of metformin during pregnancy in PCOS.
Summary
The 5th International Workshop–Conference on Gestational Diabetes brought together many leading researchers and clinicians in the field of diabetes and pregnancy from around the world. Key presentations consisted of the state-of-the art information on key topic areas of GDM. Speakers reviewed the pathophysiology and epidemiology of GDM, approaches to treatment and long-term implications for mother and child, as well as potential topics for future research. A poster session was held in conjunction with the conference to provide additional information on the key topics. As is the customary format for this conference, delegates were divided into smaller discussion groups to draft recommendations based on the research and present to all conference delegates at the final session of the conference. A complete report of these recommendations will be published in late 2006.
Future perspective
The material presented at the GDM conference and the discussions that followed suggest several key developments that are likely in the future. These include:
After the results of the HAPO study are available, new criteria for the diagnosis of GDM will be developed that are based on perinatal outcome
Recently published results and ongoing studies will support the conclusion that treatment of ‘mild GDM’ can improve perinatal outcome
The safety of the use of glyburide and metformin during pregnancy will be more definitively evaluated in long-term follow-up studies of offspring
Comparisons of the effectiveness of current and/or new therapies for GDM will take into account other hormonal and metabolic factors in addition to restoring normal blood glucose concentrations
Highlights
Gestational diabetes mellitus (GDM) is defined as ‘glucose intolerance with onset or first recognition during pregnancy’.
The prevalence of GDM has increased by 40–50% in the last decade. This increase coincides temporally with the epidemic of obesity that is present in the USA.
GDM affects 5–8% of all pregnancies in the USA, where the risk of developing Type 2 diabetes in the 5–10 years following the pregnancy is 50% or more, with a lifetime risk of approximately 80%.
In many patients, medical therapy with glyburide to augment medical nutrition therapy may achieve blood glucose control in the management of GDM similar to that seen with insulin. Additional data on perinatal outcome and long-term follow-up studies of offspring are needed.
The risk of postpartum progression to Type 2 diabetes after having GDM can be markedly reduced with lifestyle modifications or insulin-sensitizing medication.
GDM is associated with the metabolic syndrome, which also includes obesity, hyperlipidemia and hypertension. Each of these contributes to the risk of developing cardiovascular disease.
Among women with polycystic ovary syndrome, the prevalence rate of GDM has been reported to be as high as 20% in the first trimester of pregnancy and 30% at 32 weeks of pregnancy.