Gestational diabetes mellitus (GDM) is typically defined as glucose intolerance of variable severity with onset or first recognition during the second or third trimester of pregnancy (5). In contrast, overt or pregestational diabetes is defined by the American Diabetes Association (ADA) by a random glucose >200 mg/dL with classic signs and symptoms or a fasting glucose >125 mg/dL (see Table 29-1). Women with fasting hyperglycemia before 24 weeks probably have overt diabetes, as their pregnancy outcomes are similar to those with overt diabetes (6). Depending on the criteria, GDM complicates an estimated 2% to 9% of pregnancies, with the prevalence increasing over the past 20 years, most likely due to the obesity epidemic (7). The United States Preventive Services Task Force published summary recommendations in 2003 regarding population screening for GDM (8). While the task force found fair to good evidence that screening combined with diet and insulin therapy reduces fetal macrosomia, routine screening for low-risk individuals is not mandated due to insufficient evidence that it reduces important adverse health outcomes (9). Other than obesity, known risk factors for GDM include advanced maternal age, family history, glucose intolerance with prior pregnancy, and ethnicity, with women of color being at higher risk in the United States (10). Despite the lack of a mandate, survey results suggest that 96% of obstetricians universally screen during all pregnancies with an overwhelming majority (95.2%) using a 50 g glucose 1-hour oral test (11). Figure 29-1 shows a flowchart summarizing a screening and diagnostic strategy specific to GDM based on underlying risk factors. The most common diagnostic criteria used for GDM are the Carpenter–Coustan revised criteria recommended by the ADA (used by 38% of obstetricians surveyed) and the National Diabetes Data Group criteria (used by 59%) (11) (see Table 29-2). The Carpenter–Coustan criteria is more inclusive and sensitive, while still identifying patients with a higher risk for cesarean or operative vaginal delivery, macrosomia, and shoulder dystocia (12). This increased sensitivity is potentially very important beyond the perinatal period as GDM likely represents a stage in the evolution of diabetes with most, but not all, women going on to develop diabetes outside of pregnancy (13,14).

An additional 1% of pregnancies are complicated by pregestational or preexisting DM, with type 2 DM being more common than type 1 (15). As caloric and insulin needs increase during pregnancy, these women need careful monitoring to prevent both fetal morbidity and worsening of end-organ disease. Perinatal outcomes have improved markedly in recent years and are optimal when vascular disease is not present and glucose control is achieved before conception (15). Historically, White’s classification system was used to predict perinatal risk, such as prematurity or hypertensive disorders, based primarily on age of onset of diabetes and end-organ involvement. Variations of this classification scheme labeled individuals with a letter, with category A representing gestational diabetes and B–H indicating increasing duration of disease (B less than 10 years, C greater than 10 years) and presence of benign retinopathy (D), nephropathy (F), proliferative retinopathy (R), and cardiac disease (H). White’s system was used extensively from 1978 through approximately 1994 when the American College/Congress of Obstetricians and Gynecologists (ACOG) decided to shift clinical focus on whether diabetes existed before pregnancy and the adequacy of metabolic control (16).

Normal pregnancy is characterized by diurnal changes in plasma glucose and insulin with mild fasting hypoglycemia and postprandial hyperglycemia. This normal pregnancy physiology has been described as a diabetogenic state marked overall by increased insulin resistance and reduced sensitivity to insulin action (16,17). This resistance begins midway through pregnancy and peaks in the third trimester to approximate type 2 diabetes (13). The exact mechanism of insulin resistance is not fully known, but the effect is likely mediated by progesterone and estrogen, either directly or indirectly. Placental hormones, such as human placental lactogen, may also increase lipolysis, thus increasing circulating free fatty acids, and worsening tissue insulin resistance (18,19). The resultant prolonged postprandial hyperglycemia from peripheral insulin resistance likely serves to ensure a glucose supply to the fetus. Other changes to maternal glucose
homeostasis include transient hypoglycemia between meals and at night due to continuous fetal draw with mean fasting glucose levels as low as 56 mg/dL in healthy patients 28 to 38 weeks’ gestation (20). The hypothesis of the placenta as the critical endocrine organ is supported by rapid improvement in insulin resistance after delivery in normal pregnancies (13).

Women with GDM have an inadequate endogenous insulin supply to meet tissue demands during the period of increased resistance associated with pregnancy. For these women, hyperglycemia is a result of both inadequate insulin production combined with a more chronic form of insulin resistance exacerbated by physiologic changes of pregnancy (21). Beta cell dysfunction is apparent prior to conception and does not abate after delivery as in normal pregnancies, reinforcing the idea of chronic glucose intolerance unmasked by gestational changes. Women with GDM still increase insulin production in response to decreased sensitivity. However, they secrete 40% to 70% less insulin for any degree of insulin resistance, reflecting progressive loss of beta cell compensation due to a variety of mechanisms such as obesity, genetics, and autoimmune changes (13). Variations in the etiology of glucose intolerance may explain differences in severity, responses to medication, and progression of disease after pregnancy.

Glucose crosses the placenta by facilitated diffusion and leads to fetal hyperglycemia in poorly controlled diabetic states (15). If this hyperglycemia occurs early in gestation, as in pregestational diabetes, fetal anomalies (rather than chromosomal) are possible. Significant hyperglycemia during organogenesis of 5 to 8 weeks’ gestation may cause severe congenital malformations and subsequent potential for spontaneous abortion, including complex cardiac defects, central nervous system anomalies, and skeletal malformations (22). Table 29-3 shows common congenital anomalies and their incidence in infants of women with overt diabetes (22,23). Glycosylated hemoglobin levels are consistently related to the frequency of anomalies (24). The mechanism causing these congenital anomalies is not fully understood, but hyperglycemia may induce oxidative stress in the fetus which disrupts the cardiac neural crest migration and causes outflow tract defects (25). Women with preexisting DM should have preconception counseling and optimized glucose control to prevent this leading cause of perinatal mortality, with 6% to 12% of infants of women with diabetes affected by major congenital anomalies (15). Furthermore, maternal obesity itself may also be associated with an increased risk of certain types of congenital malformations, making the overall risk of an anomaly even higher when obesity and diabetes coexist
(25,26). Women with GDM alone do not appear to be at risk for fetal congenital anomalies, as hyperglycemia is probably not severe enough to impair organogenesis during that time period (27).

If the fetus survives this initial period of organogenesis, the fetal pancreatic beta cells secrete insulin in response to the abnormally high glucose load. Insulin is a potent growth hormone–stimulating excessive fetal growth, particularly in adipose tissue (15). The concept of hyperglycemia leading to fetal hyperinsulinemia and adiposity is often referred to as the Pederson hypothesis (28). Macrosomia, often defined as birth weight greater than 4,500 g, is the most commonly encountered adverse outcome in term infants of pregnancy complicated by diabetes, with a large-for-gestational-age rate of 45.2% compared with 12.6% in one population-based study (22,29). The significant increase in adipose tissue is disproportionately concentrated around the shoulders and chest, more than doubling the risk of shoulder dystocia or birth trauma at vaginal delivery, as well as increasing the rate of cesarean delivery (29,30). Figure 29-2 illustrates the excessive fetal growth that can occur with poorly controlled diabetes. The multicenter prospective Hyperglycemia and Adverse Pregnancy Outcomes (HAPO) study demonstrated continuous linear relationships between increasing maternal glucose measures and birth weight, primary cesarean delivery, clinical neonatal hypoglycemia, premature delivery, shoulder dystocia or birth injury, and preeclampsia (31). Maternal prepregnancy weight is a confounding factor in diagnosing gestational diabetes and likely is an independent risk factor for macrosomia. Unfortunately, diabetes and maternal weight are not sufficient predictors of shoulder dystocia to warrant the risk of planned cesarean delivery in all cases. Shoulder dystocia can also occur unpredictably in infants of normal birth weight.

Pregnancies complicated by diabetes are also more likely to result in prematurity or growth-restricted infants. For example, spontaneous preterm labor occurs up to 2 to 5 times more often in women with pregestational diabetes as compared to non-diabetic pregnant patients, perhaps related to increased incidence of hydramnios from poor glycemic control, fetal hyperglycemia, and polyuria (15,22). Preexisting renal dysfunction (creatinine >1.5 mg/dL) has been associated with delivery before 32 weeks’ gestation, very low birth weight, and increased incidence of neonatal hypoglycemia, independent of degree of proteinuria and glycemic control during any trimester (32). Vasculopathy, either from chronic diabetes or preeclampsia, can result in uteroplacental insufficiency and fetal growth restriction. Ultimately, these infants may be delivered early for fetal or maternal reasons, such as preeclampsia.

In addition to gestational size differences, infants of diabetic mothers are more at risk for perinatal death and stillbirth, typically associated with large-for-gestational age fetuses during the last 4 to 6 weeks of gestation (22,23). Hypothetically, fetal demise may result from villous edema, induced osmotically by hyperglycemia, leading to poor fetal oxygen transport and placental dysfunction (33). Unfortunately, early delivery is not an acceptable strategy to prevent intrauterine fetal death as these infants are also at higher risk for respiratory distress (14). After delivery, infants must be monitored carefully to avoid profound hypoglycemia. Hyperplasia of the fetal beta-islet cells in response to the maternal glucose load during gestation leads to increased circulating fetal insulin and subsequent hypoglycemia in newborn. The hypoglycemia after delivery may be related more to maternal hyperglycemia during labor, rather than reflecting chronic levels as measured by HbA_1c. Various fetal and neonatal consequences of maternal diabetes are summarized in Table 29-4.

The effects of maternal hyperglycemia do not end in the peripartum period. Prospective studies have examined the role of diabetes exposure on childhood obesity and offspring risk for type 2 diabetes. Close and long-term follow-up of the offspring of two populations, a Chicago cohort and a Pima Indian group, demonstrated increased weight and impaired glucose tolerance or prevalence of type 2 diabetes (4). This predisposition to glucose intolerance exists even after adjusting for presence of diabetes in the father and obesity in the offspring, implicating the non-genetic effect of intrauterine environment. These long-term effects appear to be similar regardless of maternal diabetes type (4). Future research in this area will need to focus on whether glycemic control can prevent the vicious cycle of obesity and diabetes.

Many maternal consequences of diabetic pregnancies are likely related to severity of preexisting disease and degree of glycemic control (15). Excessive fetal growth clearly puts the mother at higher risk for birth trauma and operative delivery and potential for associated wound infections. Infections during pregnancy, including wound infections after cesarean delivery, are more common in women with pregestational diabetes when compared to non-diabetic controls (34,35). Fortunately, the current practice of routine antimicrobial prophylaxis has resulted in low rates of wound infection and endometritis, reportedly only 0.7% and 3% in an analysis of over 200 nulliparous women with type 1 diabetes (36). Other,
more prevalent maternal complications associated with diabetes are more likely to result in a premature delivery. The risk of preeclampsia, for example, varies based on White’s classification scheme, with progressive indications of underlying vascular pathology and hypertension greatly increasing the incidence of disease as shown in Figure 29-3 (15,31). Preexisting nephropathy is associated with preeclampsia rates of 50% compared to 15% to 20% in women without renal dysfunction (14). The combination of nephropathy and hypertension is particularly predictive of growth restriction and premature delivery (15).

As demonstrated by the HAPO study, diabetes, even in milder forms, undoubtedly impacts pregnancy outcomes (31). The extent to which pregnancy influences the progression of long-term outcomes of diabetes is less clearly defined. As discussed previously, a majority of women with GDM will go on to develop overt diabetes. For women with pregestational DM, the degree of deterioration of end-organ disease as a result of pregnancy changes may depend on baseline function. Strict glycemic control in pregnancy may actually contribute to the acute progression of preexisting retinopathy (15). Although controversial, mild to moderate diabetic nephropathy probably does not progress as a result of pregnancy (15). However, women with baseline serum creatinine over 1.4 mg/dL are at risk for developing end-stage renal disease (ESRD) postpartum. In a retrospective analysis of diabetic parturients with moderate-to-severe renal dysfunction at pregnancy onset, 45% required dialysis approximately 36 months earlier than predicted by prepregnancy estimates based on linear decline in glomerular filtration rates (37). This accelerated rate of progression may be due to increased intraglomerular
pressure associated with pregnancy, exacerbation of hypertension from preeclampsia, increased incidence of urinary tract infections, or inability to use angiotensin-converting enzyme (ACE) inhibitors during gestation (37). Preeclampsia and preterm births in pregestational diabetes may also predict a long-term increased risk of ESRD and death, perhaps representing a marker for severity of disease or vasculopathy (38). Cardiovascular disease in pregestational diabetics, previously referred to as Class H, is not well studied due to limited sample sizes. These patients do appear to be at risk for myocardial infarction and death associated with the hemodynamic changes of pregnancy and the peripartum period. This mortality may be decreased in women who have undergone coronary artery bypass grafting prior to pregnancy, but definitive conclusions are not possible given limited data (39). ACOG recommends early comprehensive eye examinations and baseline evaluation of renal function by serum creatinine and urinary protein excretion for women with pregestational DM (15). Electrocardiogram and echocardiography should be considered in women with signs or symptoms of coronary artery disease (14). Information on the interaction between pregnancy and somatic or autonomic neuropathy in diabetes is limited, but the natural course of existing neuropathy is likely not substantially changed (39). Nausea and vomiting during pregnancy might be worsened in the setting of gastroparesis from autonomic neuropathy, further complicating diet and glycemic control (14).

Intensive treatment of severe hyperglycemia during pregnancy results in a reduction in the incidence of macrosomia based on summary of evidences from randomized controlled trials (9). The benefit for tight treatment of milder hyperglycemia is less clear. However, a recent randomized trial showed definite significant benefit for women with mild to moderate GDM randomized to interventions of individualized dietary changes, daily monitoring of glucose levels, and insulin therapy as needed compared to a routine-care group (40). Specifically, significant differences were found in percent with any serious perinatal complication (1% vs. 4%), birth weight, and percent with macrosomia (10% vs. 21%) (40). The benefit of treatment for mild hyperglycemia was confirmed in a subsequent randomized multicenter trial for women with mild GDM comparing outcomes of birth weight, cesarean delivery, shoulder dystocia, and rates of preeclampsia (41).

The goals of glycemic control in diabetic parturients have not been well studied and are typically based on normative values for non-diabetics during pregnancy. “Upper boundary” treatment targets are probably sufficient given that observational studies have shown an increased likelihood of small-for-gestational age infants with low mean capillary glucose levels (<87 mg/dL) (21). For example, ACOG recommends a glucose target of ≤95 mg/dL during fasting, ≤100 mg/dL preprandial, ≤140 mg/dL 1 hour after eating, and ≤120 mg/dL 2 hours after a meal (15). Trials specifically addressing whether these glycemic-control targets are appropriate for diabetics in pregnancy are needed.

With pregestational diabetes, obstetric management begins with preconception counseling, education and evaluation on diet, exercise, and insulin therapy, and nutrition and folate supplementation (15). Ideally, women who anticipate pregnancy should also have optimized control, including monitoring of pre and postprandial glucose with subsequent adjustments of insulin requirements. The adequacy of chronic metabolic control should be assessed with an HbA_1c. Select women with longstanding diabetes may need even further evaluation, including retinal examination, 24-hour urine collection for protein excretion and creatinine clearance, and electrocardiography. Thyroid function studies are also recommended in type 1 diabetics due to the high percentage of women with concomitant disease (40%) (15). The primary goal of this preconception optimization is to reduce risk of neural tube defects and congenital anomalies. Parturients without prenatal counseling have been shown to have 4 times as many fetal and neonatal death or congenital abnormalities compared to individuals with counseling (42,43).

After conception and during the first trimester, clinicians should encourage frequent self-monitoring of glucose levels with appropriate adjustments in insulin and diet. Insulin requirements may actually decrease by 10% to 20% in the first trimester, including a risk of hypoglycemia at night after prolonged fasting due to continuous fetal uptake of glucose (44). If glycemic control is poor, then hospitalization may be needed to achieve better glycemic control during this critical period of organogenesis. Historically, insulin is the mainstay of treatment for pregestational DM or poorly controlled GDM (15,21). Biosynthetic human insulin is most commonly used during pregnancy in an effort to decrease fetal antibody response to the small amount of maternal insulin that crosses the placenta bound to the IgG antibody (15,21). The pharmacologic profiles of commonly used insulins are listed in Table 29-5 (45). Insulin demands increase throughout pregnancy, so careful monitoring is a necessity to prevent negative effects of hypoglycemia. Gabbe and Graves described one strategy for initiating insulin therapy based on patient weight. The total insulin dose can be approximated as 0.8 Units per kilogram per day (U/kg/day) in the first trimester, 1.0 U/Kg/day in the second trimester, and 1.2 U/kg/day for the third trimester. Two-thirds of the total dose should be intermediate-acting (NPH or Lente, half given before breakfast the other half before bedtime), and one-third should be short-acting administered with each meal (lispro or regular, 15 or 30 minutes before eating) (14). Alternatively, subcutaneous insulin infusion therapy may be used to closely mimic physiologic insulin secretion, with approximately 50% administered basally and 50% divided before meals and snacks. Retrospective review and survey data suggests high maternal satisfaction based on continued pump use after pregnancy, but potentially higher costs of care compared to multiple insulin injections (46).

In the second and third trimesters, euglycemia remains the goal, and insulin requirements often increase along with
insulin resistance from hormonal changes. Ultrasound and alpha fetal protein can be obtained during this time period to further evaluate for potential neural tube defects and other anomalies. Ultrasound assessment of fetal abdominal circumference (AC) in the second and third trimesters for women with GDM may aid in selecting targets and intensity of therapy. Compared to conventional therapy, multiple studies have shown a reduction in large-for-gestational age infants when insulin and stricter glucose control was instituted for women with “high-risk” fetal AC above the seventy-fifth percentile (47). Other measures of glucose control, such as glycosylated hemoglobin, have not yet demonstrated value in influencing management decisions and predicting macrosomia (48).

For GDM patients, the cornerstone of treatment is medical nutrition therapy and lifestyle interventions. The food plan, ideally prescribed by a registered dietician, should meet nutrient requirements for pregnancy and restrict carbohydrate load while avoiding starvation ketosis associated with severe calorie restrictions (21,49). Nutrition practice guidelines have been shown to reduce the need for insulin compared to usual nutrition care (49). Nutrition therapy is likely to be particularly important for obese women who are prone to larger infants irrespective of diabetic status. Maternal weight gain in the first trimester has been shown to be more predictive of infant weight than gain later in pregnancy (50). The ADA Clinical Practice Recommendations suggest a moderate 30% calorie restriction for obese women (BMI >30 kg/m²) with GDM to control weight gain and glucose levels while avoiding ketosis (50).

Patients who are not adequately controlled with nutritional management or who exhibit excessive fetal growth should receive pharmacologic intervention, most commonly insulin, but more recently with oral antidiabetic agents. The three main classifications of oral pharmacologic interventions for diabetes are insulin secretagogues, insulin sensitizers, and alpha-glucosidase inhibitors. Insulin secretagogues stimulate beta cells to secrete insulin, so residual beta cell function is necessary. This class includes sulfonylureas and meglitinide, of which only glyburide has been demonstrated to have minimal placental transfer without excess neonatal hypoglycemia (21,51). Its onset of action is approximately 4 hours with a duration of 10 hours (14). Glyburide may be more beneficial in women with normal or slightly increased body weight (51). Metformin is the most commonly used insulin sensitizer, although the majority of its use in pregnancy is in women with polycystic ovarian syndrome (PCOS) (15). Metformin does cross the placenta, and at this time, beneficial or deleterious effects to the fetus are not fully known. In a prospective randomized trial comparing metformin with insulin therapy, the rate of neonatal complications based on a composite measure was not different, although severe hypoglycemia occurred more often in the insulin group. Women in the metformin group were much more likely to prefer that regimen for a subsequent pregnancy compared to insulin, but 46.3% in that group required supplemental insulin to meet glycemic targets (52). Acarbose, the alpha-glucosidase inhibitor, has also not been studied extensively, but preliminary results suggest reduced postprandial glucose in GDM with expected abdominal cramping (21). In a recent systematic review of the literature comparing insulin with all oral hypoglycemic agents, only four randomized controlled trials and five cohort studies were identified that had appropriate diagnostic criteria and comparison groups for maternal and fetal outcomes (53). No significant differences were found in maternal glycemic control or cesarean delivery rates with similar infant birth weights among women treated with either insulin or glyburide. Neonatal hypoglycemia was more common (8.1%) with insulin when compared with metformin (3.3%). Rates of congenital malformations did not differ in infants of women treated with oral agents versus insulin, suggesting that if placental transfer occurs, the impact on the fetus is neutral or at least not harmful (53).

Despite aggressive therapy, the physiologic changes associated with pregnancy may contribute to the development of diabetic ketoacidosis (DKA) in 5% to 10% of pregnancies with pregestational DM. DKA is more common in type 1 diabetics and occurs with more frequency during pregnancy due to worsening insulin resistance (15). The pathophysiology of DKA is summarized in Figure 29-4 (54). This life-threatening emergency can develop rapidly during pregnancy and with less extreme hyperglycemia (55). Case reports have even described “euglycemic” DKA during pregnancy with initial presentation of nausea, abdominal pain, ketonuria and high anion gap metabolic acidosis but with a normal glucose of 77 mg/dL. The parturient improved appropriately with insulin and dextrose infusions (55). Risk factors for the development of DKA include new onset during pregnancy, infections, poor patient compliance, insulin pump malfunction, and treatment with beta-mimetic tocolytic medications or antenatal corticosteroids (15). DKA can occur during pregnancy without precipitating events other than emesis. In a small case series of 37 parturients with DKA, 42% had emesis with rapidly evolving starvation ketosis and no known precipitating factors (56). The management strategy for DKA during pregnancy is described in Table 29-6 and typically involves intensive care unit monitoring for aggressive hydration, insulin infusion, and frequent assessment of glucose and potassium concentrations (15). Continuous fetal monitoring may show recurrent late decelerations that improve with maternal condition. The fetal mortality rate has improved recently from 35% to approximately 10% of cases (39).

Rather than DKA often associated with type I diabetes, pregnant patients with pregestational type 2 diabetes may be more prone to develop a hyperosmolar hyperglycemic non-ketotic state (HHNS) (57). HHNS is characterized by hyperglycemia, hyperosmolality (often >360 mOsm/L), and extreme hypovolemia without ketonemia. Patients may have mental status changes including confusion, somnolence, and possible coma or seizure activity as hyperosmolarity increases. At least initially, the syndrome occurs without ketosis or acidosis, unless superimposed with other metabolic acidoses, such as infection, sepsis, dehydration-related renal failure, or lactic acidosis. At this time, HHNS outcomes in pregnancy are limited to case reports but the incidence may increase along with the prevalence of type 2 diabetes associated with the obesity epidemic. The hallmark of treatment is volume resuscitation with metabolic derangements corrected as appropriate (57,58). The average fluid deficit in non-pregnant patients is approximately 9 L. Relatively small-dose insulin infusions are usually used to correct hyperglycemia after volume replacement has been initiated. Metabolic, electrolyte, and fluid abnormalities may put the parturient at risk for intrauterine fetal demise (IUFD), as occurred in at least two published cases (57,58). Overly rapid correction of maternal glucose can also cause adverse fluid and osmotic events in the placenta. Placental perfusion may also be compromised by the overall dehydration and blood volume reduction caused by osmotic diuresis from sustained glycosuria (57). Finally, HHNS heightens the risk for thromboembolic events during pregnancy, presumably due to stasis
from the low blood volume state. Prophylactic heparin may be indicated in situations where the patient is remote from delivery (57).

In addition to the increased incidence of hyperglycemic crises, women with pregnancies complicated by diabetes are also prone to hypoglycemia. Patients and families should be taught how to respond quickly and appropriately to signs and symptoms of hypoglycemia (often defined as glucose less than 60 mg/dL), with ACOG recommending a glass of milk over fruit juice. Type I diabetics may also need glucagon on hand for severe hypoglycemia and loss of consciousness (15). During labor and delivery, maternal glucose should be kept in the high-normal range (approximately 100 mg/dL) to prevent the extremes. ACOG’s recommendations for insulin management during labor and delivery are shown in Table 29-7 (15). Hypoglycemia should be managed with the understanding that overshooting maternal targets may increase the incidence of fetal hypoglycemia after delivery. If glucose levels are less than 70 mg/dL and oral intake is not possible as in the case of scheduled elective cesarean delivery, 2 to 5 g of glucose can be administered intravenously, with appropriate communication to the neonatal team regarding need for maternal glucose administration (59).

Preterm labor in diabetic parturients should be managed carefully with close monitoring of maternal glucose levels. Beta-adrenergic agents, such as terbutaline, may cause hyperglycemia, making magnesium the tocolytic of choice. Antenatal corticosteroids to promote fetal lung maturity will also complicate management, with increased insulin requirements expected over the next 5 days after administration (15). A consideration of fetal and maternal conditions is necessary to determine optimal timing of delivery (15).

Published recommendations from the Fifth International Workshop-Conference on GDM held in 2005 do not support routine delivery before 38 weeks’ gestation without evidence of specific maternal or fetal compromise (21). Mode and time of delivery has to be assessed based on multiple factors such as severity and control of diabetes, previous obstetric
history, cervix favorability, fetal size, fetal and maternal well-being, and co-morbidities such as preeclampsia (60,61). Similarly, amniocentesis to determine fetal lung maturity is not indicated for well-controlled patients with indications for delivery in well-dated pregnancies (21). Presumably, the indication for delivery would supercede the need for establishment of fetal lung maturity prior to 38 weeks’ gestation. Other types of fetal assessment, such as non-stress and contraction stress tests and Doppler umbilical artery flow velocimetry, are now more commonly performed to assess or confirm fetal well-being after 32 weeks’ gestation and assist with delivery decisions (62,63,64). The diagnosis of GDM by itself does not indicate a need for cesarean delivery to prevent birth trauma. Fetal surveillance techniques are not yet able to detect fetal body asymmetry and predict risk of shoulder dystocia (60). However, planned cesarean delivery may be considered in extreme circumstances, such as estimated fetal weights greater than 4,500 g (15). The actual cesarean rate may be as high as 50% to 80% among overt diabetic parturients as the chance of successful labor induction decreases with the severity of diabetes (61,65).