INTRODUCTION — Treatment of gestational diabetes mellitus (GDM) can improve pregnancy outcome. Many patients can achieve glucose target levels with nutritional therapy alone, but up to 30 percent will require pharmacotherapy [1]. Even patients with mildly elevated glucose levels who do not meet standard criteria for GDM may have better pregnancy outcomes if treated as GDM since the relationship between glucose levels and adverse pregnancy outcomes such as macrosomia exists continuously across the spectrum of increasing glucose levels [2-9].

The general approach to glucose management in patients with GDM is reviewed here. Screening, diagnosis, and obstetric management are discussed separately. (See "Gestational diabetes mellitus: Screening, diagnosis, and prevention" and "Pregestational (preexisting) diabetes mellitus: Obstetric issues and management".)

RATIONALE FOR TREATMENT — Treatment of GDM is important to minimize maternal and neonatal morbidity. In a U S Preventive Services Task Force meta-analysis of randomized trials, compared with no treatment, appropriate management of GDM (nutritional therapy, self-blood glucose monitoring, administration of insulin when target blood glucose concentrations are not met with diet alone) resulted in reductions in [10]:

●Preeclampsia (7.2 versus 11.7 percent; relative risk [RR] 0.62, 95% CI 0.43-0.89; three trials, moderate quality)

●Birth weight >4000 g (8.4 versus 17.4 percent; RR 0.50, 95% CI 0.35-0.71; five trials, moderate quality)

●Shoulder dystocia (1.5 versus 3.5 percent; RR 0.42, 95% CI 0.23-0.77; three trials, moderate quality)

The only identified potential harm or cost resulting from treatment of GDM was an increased number of prenatal visits; however, the frequency and consequences of maternal hypoglycemia in patients treated with insulin were not reported. No statistically significant differences in rates of cesarean birth, induction of labor, small for gestational age newborns, neonatal hypoglycemia, neonatal hyperbilirubinemia, neonatal respiratory complications, birth trauma, or neonatal intensive care unit admission were demonstrated compared with no treatment, although the quality of evidence was low.

Some authors have suggested that addressing maternal obesity and excessive gestational weight gain may be more important than detecting and treating GDM because maternal weight may be more closely related to adverse outcomes, particularly fetal overgrowth, than glucose intolerance [11]. However, data from the Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study refute this hypothesis. In HAPO, both obesity and GDM (International Association of Diabetes and Pregnancy Study Groups criteria) were independently predictive of macrosomia, preeclampsia, primary cesarean birth, and neonatal adiposity [2].

Few studies have evaluated long-term effects of maternal treatment for GDM on offspring. In one study of treated versus untreated patients with mild GDM (defined as fasting plasma glucose <95 mg/dL and two of three timed measurements that exceeded established thresholds [1 hour, 180 mg/dL; two hours, 155 mg/dL; three hours, 140 mg/dL]), treatment during pregnancy did not reduce late adverse metabolic outcomes (eg, obesity, glucose intolerance) in offspring age 5 to 10 years [12]. This may reflect lack of a true treatment effect, inadequate treatment of hyperglycemia during pregnancy, the mildness of the glucose intolerance, or inadequate power to show modest differences in outcome because of the low prevalence of these disorders prior to puberty, and the small numbers of study participants.

MEDICAL NUTRITIONAL THERAPY — Medical nutritional therapy is the process by which the dietary plan is tailored for patients with diabetes, based on medical, lifestyle, and personal factors. Patients with GDM should receive medical nutritional counseling by a registered dietitian (when possible) upon diagnosis and be placed on an appropriate diet. The goals are to [13]:

●Achieve normoglycemia

●Prevent ketosis

●Provide adequate nutrition

●Provide adequate gestational weight gain based on maternal body mass index (BMI)

●Contribute to fetal well-being

Most patients (up to 85 percent) with GDM based on Carpenter and Coustan criteria can achieve target glucose levels with lifestyle modification alone [3,14]. A detailed review of medical nutritional treatment of individuals with diabetes can be found separately. (See "Nutritional considerations in type 1 diabetes mellitus".)

The specific diet for patients with GDM that achieves optimum maternal and newborn outcomes is unclear [15-17]. A key simple, achievable intervention is to emphasize the benefits of elimination, or at least reduction, of consumption of sugar-sweetened beverages (eg, soft drinks, fruit drinks) and encourage drinking water instead. Noncaloric sweeteners may be used in moderation. Traditionally, restricting carbohydrate intake (particularly simple carbohydrates) has been favored because it reduces postprandial hyperglycemia [18] and fetal overgrowth [19,20] despite the lack of high-quality evidence. In a systematic review of randomized trials comparing a variety of dietary interventions (eg, low glycemic index, DASH, low carbohydrate, energy restriction, soy protein, fat modification, ethnic, high fiber) with conventional dietary recommendations for patients with GDM (18 trials, 1151 participants), dietary intervention overall reduced fasting and postprandial glucose levels (fasting: -4.07 mg/dL, 95% CI -7.58 to -0.57; postprandial -7.78 mg/dL, 95% CI -12.27 to -3.29), need for medication (relative risk [RR] 0.65, 95% CI 0.47-0.88), birth weight (-170.62 g, 95% CI -333.64 to -7.60), and macrosomia (RR 0.49, 95% CI 0.27-0.88) [16]. When analyzed by diet subtype, low glycemic index, DASH, low carbohydrate, and ethnicity-based diets had beneficial effects on maternal glucose levels. A limitation of the analysis was that all of the trials had small sample sizes.

Probiotics and high fiber diets do not appear to improve glycemic control [21,22].

Meal plan — A typical meal plan for patients with GDM includes three small- to moderate-sized meals and two to four snacks. Adjustment of the meal plan should be ongoing and based upon results of self-glucose monitoring, appetite, and weight-gain patterns, as well as consideration for maternal dietary preferences and work, leisure, and exercise schedules.

Close follow-up is important to ensure nutritional adequacy. Individual assessment and self-blood glucose monitoring are used to determine and modify specific nutrition/food recommendations. If insulin therapy is added to nutritional therapy, a primary goal is to maintain carbohydrate consistency at meals and snacks to facilitate insulin adjustments.

Calories — There are no clear data suggesting that the caloric requirements of pregnant patients with GDM are different from those without GDM [23]. For individuals with a prepregnancy BMI in the healthy range, caloric requirements in the first trimester are the same as before pregnancy and generally increase by 340 calories per day in the second trimester and 452 calories per day in the third trimester [24]. Individuals who are underweight, overweight, or obese should work with a registered dietician to determine their specific caloric requirements. (See "Gestational weight gain".)

Carbohydrate intake — Once the caloric needs are calculated, carbohydrate intake is determined as it is the primary nutrient affecting postprandial glucose levels. The total amount of carbohydrate, the distribution of carbohydrate over meals and snacks, and the type of carbohydrate can be manipulated to blunt postprandial hyperglycemia. However, reducing carbohydrates to decrease postprandial glucose levels may lead to higher consumption of fat, which may have adverse effects on maternal insulin resistance and fetal body composition.

Dietary Reference Intakes (DRI) for all pregnant people is a minimum of 175 g of carbohydrate per day and 28 g of fiber [23]. There is sparse evidence from randomized trials as to the ideal carbohydrate intake for treatment of GDM. We limit carbohydrate intake to 40 percent of total calories while ensuring that ketonuria does not ensue [25,26]. Larger, adequately powered studies are needed to evaluate the effect of various dietary interventions on perinatal outcomes in GDM.

Many patients will need individual adjustment of the amount of carbohydrate by 15 to 30 g at each meal, depending on their postprandial glucose levels, which are directly dependent upon the carbohydrate content of the meal or snack [27]. The postprandial glucose rise can be blunted if the diet is carbohydrate restricted.

In a meta-analysis of randomized trials of dietary intervention in patients with GDM, low carbohydrate diets had a favorable effect on postprandial blood glucose concentrations and significantly lowered the need for insulin therapy but did not affect other maternal or newborn outcomes (eg, macrosomia, cesarean birth, gestational weight gain), although the data were insufficient to detect small or moderate statistical differences in obstetric outcomes between the patient groups [28]. (See "Dietary carbohydrates", section on 'Glycemic index' and "Nutritional considerations in type 2 diabetes mellitus", section on 'Glycemic index and glycemic load'.)

Protein and fat intake — The remaining calories come from protein (20 percent of total calories or approximately 71 g per day [23]) and fats (40 percent of total calories; saturated fat intake should be <7 percent of total calories). Protein intake should be distributed throughout the day and included in all meals and snacks to promote satiety, slow the absorption of carbohydrates into the bloodstream, and provide adequate calories. A bedtime high-protein snack is recommended to prevent accelerated (ie, starvation) ketosis overnight and maintain fasting glucose levels within the target range.

Gestational weight gain/loss — After prescribing the diet, it is important to pay attention to subsequent changes in weight. In a retrospective cohort study including over 31,000 patients with GDM, those with appropriate gestational weight gain (table 1) had optimal outcomes, while excessive gestational weight gain was associated with a significantly increased risk of having a large for gestational age newborn, preterm birth, and cesarean birth [29]. Although suboptimal weight gain increased the likelihood of avoiding pharmacotherapy of GDM and decreased the likelihood of having a large for gestational age newborn, there were more small for gestational age newborns in this group (7.3 versus 5.6 percent). The data in this study were not corrected for potential confounders, such as smoking. (See "Obesity in pregnancy: Complications and maternal management" and "Gestational weight gain", section on 'Recommendations for gestational weight gain'.)

Some patients experience a minimal (one to five pounds) weight loss or weight stabilization for the first few weeks after starting nutritional therapy, which should be evaluated in the overall context of gestational weight gain and ongoing surveillance of weight gain in the weeks thereafter. Weight loss is generally not recommended during pregnancy, although controversy exists regarding this recommendation for patients with obesity, especially class II or III. For pregnant people with obesity, a modest energy restriction of 30 percent below the DRI for pregnant people (175 g carbohydrate, 71 g protein, 28 g fiber [23]) can often be achieved while meeting gestational weight gain recommendations and without causing ketosis [30]. (See "Gestational weight gain", section on 'Recommendations for gestational weight gain'.)

EXERCISE — Adults with diabetes are encouraged to perform 30 to 60 minutes of moderate-intensity aerobic activity (40 to 60 percent maximal oxygen uptake [VO₂max]) on most days of the week (at least 150 minutes of moderate-intensity aerobic exercise per week). A program of moderate exercise is recommended as part of the treatment plan for patients with diabetes as long as they have no medical or obstetric contraindications to this level of physical activity. Exercise that increases muscle mass, including circuit training, appears to improve glycemic control, primarily from increased tissue sensitivity to insulin. As a result, exercise can reduce both fasting and postprandial blood glucose concentrations and, in some patients with GDM, the need for insulin may be obviated [31]. (See "Exercise during pregnancy and the postpartum period" and "Effects of exercise in adults with diabetes mellitus".)

GLUCOSE MONITORING — Patients should self-monitor their glucose concentrations. Glucose meters measure capillary blood glucose, almost all available glucose meters provide plasma equivalent values rather than whole-blood glucose values. Thus, results from most available glucose meters and venous plasma glucose measured in a laboratory should be comparable. (See "Glucose monitoring in the management of nonpregnant adults with diabetes mellitus".)

Intermittent self-monitoring of blood glucose — We suggest that patients self-monitor blood glucose levels [32-35]:

●Before breakfast (ie, fasting glucose level) and

●At one or at two hours after the beginning of each meal

Results should be recorded in a glucose log, along with dietary information. This facilitates recognition of glycemic patterns and helps to interpret results stored in the memory of glucose meters.

We prefer the one-hour postprandial measurement as it corresponds more closely to the maximum insulin peak in patients using rapid-acting insulin analogs. The value of fasting plus postprandial versus preprandial measurement was suggested by a trial that randomly assigned 66 insulin-treated patients with GDM to management according to results of fasting plus postprandial monitoring (one hour after meals) or according to preprandial-only blood glucose concentrations [36]. One-hour postprandial monitoring had several benefits as compared with preprandial monitoring: better glycemic management (glycated hemoglobin [A1C] value 6.5 versus 8.1 percent), a lower incidence of large for gestational age newborns (12 versus 42 percent), and a lower rate of cesarean birth for cephalopelvic disproportion (12 versus 36 percent).

Can the frequency of self-monitoring be reduced? — When initially diagnosed with GDM, we ask patients to measure their blood glucose concentration at least four times daily, as described above. Multiple daily measurements allow recognition of patients who should begin pharmacologic therapy. There is no strong evidence regarding the frequency of glucose testing, particularly in patients on nutritional therapy who consistently have glucose levels in the target range (see 'Glucose target' below) [37,38], but decreasing the frequency of testing to every other day in specific patients with mild GDM (defined as no more than intermittent glucose elevations that are 5 to 10 mg/dL above targets), no signs of fetal overgrowth (defined as abdominal circumference (AC) >75th percentile or estimated fetal weight (EFW) ≥90^th percentile), and normal amniotic fluid volume (ie, no polyhydramnios), likely improves the patient's quality of life. In a randomized trial of patients with GDM on nutritional therapy who demonstrated glucose levels in the target range after one week of four times daily glucose testing, those assigned to testing glucose levels every other day had similar birth weights and frequency of macrosomia as those who continued to test four times daily [39].

Continuous glucose monitoring — Continuous glucose monitoring (CGM) allows determination of peak postprandial glucose levels, mean glucose level, episodes of nocturnal hyperglycemia, and percent time in range for a 24-hour period. We do not routinely use CGM in patients with GDM because of cost and it has not been proven to improve maternal or fetal outcome, but few trials have been performed. When CGM was compared with frequent self-monitoring of blood glucose in a meta-analysis of two small randomized trials, outcomes were similar for both approaches: cesarean birth (risk ratio [RR] 0.91, 95% CI 0.68-1.20), large for gestational age newborn (RR 0.67, 95% CI 0.43-1.05), neonatal hypoglycemia (RR 0.79, 95% CI 0.35-1.78) [40]. There were no perinatal deaths. Larger trials may clarify whether the observed favorable trends observed are real.

Although use of CGM has no clear advantages for most patients, it may be considered in patients who cannot consistently check fingerstick glucose levels and are willing to wear a device. Cost may be a barrier to use.

Glucose target — The American Diabetes Association (ADA) and the American College of Obstetricians and Gynecologists (ACOG) recommend the following upper limits for glucose levels, with insulin therapy initiated if they are exceeded, but acknowledge that these thresholds have been extrapolated from recommendations proposed for pregnant patients with preexisting diabetes [23,41]:

●Fasting and preprandial blood glucose concentration: <95 mg/dL (5.3 mmol/L)

●One-hour postprandial blood glucose concentration: <140 mg/dL (7.8 mmol/L)

●Two-hour postprandial glucose concentration: <120 mg/dL (6.7 mmol/L)

These targets are well above the mean glucose values in pregnant people without diabetes described in a literature review of studies of the normal 24-hour glycemic profile of pregnant people [42]. In this review, which had a total of 255 pregnant people without diabetes who were mostly in the late third trimester and without obesity, the pooled weighted mean glucose values (±1 SD) were fasting 71±8 mg/dL (3.9±0.4 mmol/L), one-hour postprandial 109±13 mg/dL (6.0±0.7 mmol/L), two-hour postprandial 99±10 mg/dL (5.5±0.6 mmol/L), and 24-hour glucose 88±10 mg/dL (4.9±0.6 mmol/L). These levels were derived from measurements on whole blood, plasma, self-monitored capillary glucose measurements, or tissue fluid (CGM). Although glucose levels in whole blood, plasma, and interstitial fluid differ, there was some consistency in the results. If two standard deviations are added to the means outlined in the systematic review, the upper limit of normal fasting glucose would be 87 mg/dL (4.8 mmol/L), the corresponding one-hour postprandial value would be 135 mg/dL (7.5 mmol/L), and the upper limit of normal two-hour value would be 119 mg/dL (6.6 mmol/L); while the fasting value is somewhat lower than the target 95 mg/dL (5.3 mmol/L), the postprandial values are not dissimilar to the targets described above.

Glycated hemoglobin — A1C may be a helpful ancillary test in assessing glycemic management during pregnancy [43,44]. It is not clear whether or how often it should be monitored in patients with GDM whose glucose levels are in the target range. If measured and there is a discrepancy between the A1C and glucose values, then potential causes should be investigated. (See "Measurements of glycemic control in diabetes mellitus", section on 'Racial/ethnic differences' and "Measurements of glycemic control in diabetes mellitus", section on 'Unexpected or discordant values'.)

Good normative data for A1C in each trimester are not available. A1C values tend to be lower in pregnant compared with nonpregnant people [45] because the average blood glucose concentration is approximately 20 percent lower in pregnant people, and in the first half of pregnancy, there is a rise in red cell mass and a slight increase in red blood cell turnover [46,47]. Other factors that have been reported to affect A1C levels include race (although it is not clear whether the higher A1C levels observed in Black persons compared with White persons are due to differences in glucose levels or racial differences in the glycation of hemoglobin [48]) and iron status (chronic iron deficiency anemia increases A1C, treatment of iron deficiency anemia with iron lowers A1C). Sources of variation in A1C levels are discussed in detail separately. (See "Measurements of glycemic control in diabetes mellitus", section on 'Glycated hemoglobin (A1C)'.)

MONITORING FOR KETONURIA — We do not routinely monitor urinary ketones in pregnant people with GDM, as diabetic ketoacidosis is extremely rare in patients first diagnosed with diabetes during pregnancy [49,50]. Episodes of physiological ketonemia and ketonuria are not uncommon in pregnancy and can occur with hypocaloric diets [51]. Studies have reported inconsistent findings regarding a potential association between ketonuria and impaired cognitive outcome in offspring [52-56].

PHARMACOTHERAPY

Goal — The goal of pharmacotherapy is to manage glucose levels so that the majority are no higher than the upper limit of the target range without inducing any episodes of hypoglycemia. Overly tight metabolic control in GDM (ie, average blood glucose levels ≤86 mg/dL) has no additional benefits and increases the risk for iatrogenic growth restriction [57,58].

Indications for pharmacotherapy

●Glucose levels above the target range – If glucose targets cannot be maintained by medical nutritional therapy, then pharmacotherapy should be initiated, but the degree of hyperglycemia at which the disadvantages of initiating insulin therapy are clearly outweighed by the benefits has not been definitively determined and varies among providers [59]. We initiate pharmacotherapy when over 30 percent of the blood glucose values in a week are above target glucose thresholds (see 'Glucose target' above). In a retrospective cohort study of patients with GDM, initiating pharmacotherapy when 20 to 39 percent of glucose levels were above goal compared with ≥40 percent above goal was associated with a reductions in preterm birth (7.4 versus 15.7 percent), neonatal intensive care unit admission (4.0 versus 11.7 percent), and large for gestational age newborn (9.1 versus 21.2 percent) [60]. Our general approach is described in the algorithm (algorithm 1).

●Fetal overgrowth – Data from some randomized trials suggest that pharmacotherapy, specifically insulin, in the subgroup of patients with indirect evidence of fetal hyperinsulinemia (eg, abdominal circumference [AC] >75^th percentile or estimated fetal weight [EFW] ≥90^th percentile on the early third-trimester sonogram) can reduce the occurrence of macrosomia and large for gestational age in newborns, even in patients with GDM who are not hyperglycemic at the time pharmacotherapy is initiated. In a meta-analysis including only two trials, compared with conventional hyperglycemia-based management in patients with a broad severity spectrum of GDM, initiation of pharmacotherapy based on ultrasound findings of a large AC increased the percent of patients requiring insulin treatment (34 versus 23 percent, relative risk [RR] 1.58, 95% CI 1.14-2.20) and reduced the occurrence of large for gestational age newborns (RR 0.58, 95% CI 0.34-0.99) and macrosomia (RR 0.32, 95% CI 0.11-0.95) without increasing the risk for small for gestational age newborns [61]. Rates of pregnancy-associated hypertension and cesarean birth were similar in both groups; data on frequency of maternal hypoglycemia were not provided. Thus, it is reasonable for patients with sonographic signs of fetal overgrowth to receive insulin to decrease the risk of large for gestational age and macrosomia despite having less than 30 percent of glucose values above target threshold [38].

Choice of pharmacotherapy — There are two pharmacologic options in pregnant patients who require pharmacotherapy: insulin (and some insulin analogs) and selected oral antihyperglycemic agents (eg, metformin, glyburide).

We favor insulin because it is effective and easily adjusted based on glucose levels, and data are lacking regarding long-term outcomes in offspring exposed to oral antihyperglycemic drugs in utero. We believe that oral antihyperglycemic agents are a reasonable alternative to insulin for patients in whom pharmacotherapy is indicated but who decline to take, or are unable to comply with, insulin therapy. Our approach is generally consistent with national and international guidance [1,23,41,62,63]. Some guidelines consider oral antihyperglycemic drugs an acceptable first-line approach in select patients, such as those with normal fasting blood glucose levels and modest postprandial hyperglycemia [62-64]. (See 'Society guideline links' below.)

Meta-analyses comparing use of oral antihyperglycemic agents with insulin therapy have generally found that both approaches can improve some pregnancy outcomes in patients with GDM or type 2 diabetes [59,65-70]. There is a trend toward more frequent maternal hypoglycemia with use of insulin [68], and some patients on oral agents need supplemental insulin to achieve and maintain glucose levels in the target range [71]. However, it is difficult to draw firm conclusions about the optimal approach because of inconsistencies in criteria for GDM, glucose targets, patient adherence to treatment, clinical outcome measures across studies, and lack of long-term safety data [68].

In randomized trials, compared with insulin, metformin:

●Reduced gestational weight gain (mean difference -1.31 kg, 95% CI -2.34 to -0.27) [69]

●Reduced birth weight (mean difference -74 g, 95% CI -115 to -33) [69]

●Reduced risk for macrosomia (odds ratio [OR] 0.60, 95% CI 0.45-0.79) [69]

●Reduced risk for neonatal hypoglycemia (risk ratio [RR] 0.63, 95% CI 0.45-0.87) [70]

●Reduced risk for pregnancy-induced hypertension (RR 0.56, 95% CI 0.37-0.85) [70]

●Increased offspring body mass index (BMI, by 0.8 kg/m²) and adiposity by mid-childhood [72]

●Differences in other outcomes were not statistically significant: large for gestational age newborn (OR 0.87, 95% CI 0.66-1.14) [69], preterm birth (RR 1.18, 95% CI 0.67-2.07), small for gestational age newborn (RR 1.20, 95% CI 0.67-2.14), perinatal mortality (RR 0.82, 95% CI 0.17-3.92), cesarean birth (RR 0.97, 95% CI 0.80-1.19) [70].

Compared with insulin, glyburide:

●Increased mean birth weight (mean difference 290 g, 95% CI 68-511) [69]

●Increased risk for macrosomia (OR 1.38, 95% CI 1.01-1.89) [69]

●Increased the frequency of neonatal hypoglycemia (12.2 versus 7.2 percent; difference 5.0, 95% CI 0.5-9.5) [73]

●Showed trends toward an increased risk for a large for gestational age newborn (OR 2.49, 95% CI 0.79-7.81) and less maternal gestational weight gain (mean difference -0.68 kg, 95% CI -1.69 to 0.34 kg) [69].

Insulin

Dose — The dose of insulin required to achieve target glucose levels varies among individuals, but the majority of studies have reported a total insulin dose ranging from 0.7 to 2 units per kg (present pregnant weight). The titration of the dose to blood glucose levels is based upon frequent self-monitoring. At least four daily glucose measurements are required (fasting and one or two hours postprandial with the addition of pre-lunch and pre-dinner measurements as needed) to optimize therapy and ensure timely increases in dose as insulin requirements increase with pregnancy progression. The insulin requirement in twin gestations complicated by GDM may double with pregnancy progression.

We do not use insulin pumps in patients with GDM because there are no data to suggest that they are necessary or more effective than conventional therapy, and the cost of an insulin pump is not justified over the relatively short duration of a pregnancy. However, case reports have described successful use in some pregnant people.

Pragmatic approach to management of hyperglycemia — Hospitalization is not necessary to initiate insulin therapy; however, if teaching some patients the procedures they need to know is not possible in the outpatient setting, then an inpatient stay to utilize the expertise of the hospital's nursing staff may justify the cost of hospitalization.

One principle we have found useful is to start with the simplest regimen and increase the complexity as needed to address the particular situation. Typically, regardless of body weight, insulin dosing is based on the glucose levels recorded in the patient's blood glucose log.

The following is our general approach to management of patients diagnosed with GDM after screening at 24 to 28 weeks who have mostly postprandial hyperglycemia, fetal AC>75^th percentile, or EFW ≥90^th percentile. Because any insulin regimen requires serial dosing adjustments in response to specific fasting or postprandial glucose levels, the starting dose should be considered just that, a starting point. Weekly glucose log review is recommended so that insulin doses can be adjusted as needed to meet target glucose levels as the pregnancy advances. Some patients may be diagnosed with diabetes and therapy initiated early in pregnancy (prior 24 to 28 weeks screening); these patients are managed differently and generally require slightly lower insulin doses since insulin resistance is lower early in pregnancy.

●We begin with a single injection of 10 to 20 units of intermediate-acting basal insulin (neutral protamine Hagedorn [NPH]) and 6 to 10 units of rapid-acting insulin (lispro or aspart) in the morning immediately before breakfast; the dose within this range is based on the degree of elevation above target levels.

●If postprandial glucose levels throughout the day remain high, adjustments in the rapid-acting insulin dose are typically in the range of 10 to 20 percent. The upper end of this range is not likely to lead to hypoglycemia in patients with both obesity and GDM unless a meal is omitted after insulin is given.

●If only the post-dinner glucose level remains elevated, then we add an injection of 6 to 10 units of rapid-acting insulin immediately before dinner.

●If only the post-lunch glucose level remains elevated, we add an injection of 6 to 10 units of rapid-acting insulin immediately before lunch.

●If the fasting glucose level is elevated after postprandial levels in the target range, we add an intermediate-acting basal insulin, preferably at bedtime but with dinner is another option on an individualized basis. The initial dose is 0.2 units/kg body weight.

Dosing based on glucose levels and weight — An alternative approach to insulin therapy, somewhat more complex and likely most appropriate for individuals whose glucose levels are not well managed with simpler paradigms, is described below:

●If only the fasting blood glucose concentration is high, an intermediate-acting basal insulin (NPH) is given before bedtime but before dinner is another option on an individualized basis; an initial dose of 0.2 unit/kg body weight is utilized. A long-acting insulin analog (insulin glargine or detemir) may be used instead if NPH insulin is not available.

●If only postprandial blood glucose concentrations are high, 6 to 10 units of rapid-acting insulin analogs (aspart or lispro) are given immediately before meals. (See "General principles of insulin therapy in diabetes mellitus".)

●If both preprandial and postprandial blood glucose concentrations are high or if the patient's postprandial glucose levels can only be blunted when starvation ketosis occurs, then a four-injection-per-day regimen is utilized, which improved glycemic control and perinatal outcome compared with a twice-daily regimen in one randomized trial [74], although macrosomia rates were not impacted. The starting dose is calculated by trimester of pregnancy and body weight: 0.9 units/kg in the second trimester and 1.0 units/kg in the third trimester, split into basal and bolus dosing. In patients with class II or III obesity, the initial doses of insulin may need to be increased to 1.5 to 2 units/kg to overcome the combined insulin resistance of pregnancy and obesity.

Two-thirds of the total daily dose is administered in the morning, with two-thirds of the morning dose given as basal insulin and one-third given as rapid-acting insulin up to 15 minutes before breakfast. One-third of the total daily dose is administered in the evening, with half of this dose given as rapid-acting insulin up to 15 minutes before dinner and the other half given as basal insulin as a nighttime dose (usually at bedtime but before dinner is another option on an individualized basis). A lunchtime dose of rapid-acting insulin may be added if there is continued postprandial lunch hyperglycemia.

Management of hypoglycemia — Hypoglycemia in pregnancy is defined as a blood glucose <60 mg/dL (3.3 mmol/L). Hypoglycemia remote from meal or snack time is rare in patients with GDM treated with pharmacotherapy, and it is treated by administering 10 to 20 g of a fast-acting carbohydrate snack immediately. The American Diabetes Association (ADA) suggests the following options: 4 ounces (1/2 cup) of juice or regular soda, 8 ounces (1 cup) of skim milk, or 5 to 6 hard candies (eg, Life-Savers); glucose tablets can also be used (check package for grams per tablet as content varies). Since the sugars in milk release more slowly into the bloodstream than pure sugar options, the glucose pattern seen with pure sugars (ie, rapid elevation of glucose followed by a rapid decline) may be dampened. (See "Hypoglycemia in adults with diabetes mellitus", section on 'Reversing hypoglycemia'.)

Once patients feel better, they may check their blood glucose 15 to 30 minutes after treatment. If the glucose remains <60 mg/dL (3.3 mmol/L), repeat treatment may be necessary. On the other hand, they may need to give themselves extra insulin to compensate for overtreatment of the symptoms.

If low glucose values are encountered more than once at the same time of day, insulin doses are adjusted downward accordingly.

Type of insulin — Use of insulin preparations of low antigenicity may minimize the transplacental transport of insulin antibodies: Human insulin is the least immunogenic of the commercially available preparations. The three rapid-acting insulin analogs (lispro, aspart, glulisine) are comparable in immunogenicity to human regular insulin, but only lispro and aspart have been investigated in pregnancy and shown to have acceptable safety profiles, minimal transfer across the placenta, and no evidence of teratogenesis. Neonatal outcomes are similar to those of patients treated with regular insulin [59]. These two insulin analogs both improve postprandial excursions compared with human regular insulin and are associated with lower risk of delayed postprandial hypoglycemia.

Long-acting insulin analogs (insulin glargine, insulin detemir) have not been studied as extensively in pregnancy, but data from patients with preexisting diabetes and studies of placental transfer suggest that both detemir and glargine are safe and effective for use in pregnancy [75-81] (see "Pregestational (preexisting) diabetes mellitus: Antenatal glycemic control", section on 'Type of insulin'). Based on available data, we prefer use of human NPH insulin as part of a multiple injection regimen in pregnant people with GDM, especially given the peak at four to six hours after the morning dose, which can help decrease lunch postprandial blood glucose levels without an additional dose of rapid-acting insulin [82]. There are good data supporting the safety and effectiveness of NPH in pregnancy, and doses can be adjusted frequently and quickly in response to changing requirements in pregnant patients. If a longer-acting insulin analog is used, we prefer detemir insulin because it can be dosed twice a day, similar to NPH, with the advantage over NPH of more consistent absorption and less variability in absorption among patients. Insulin detemir is preferred over insulin glargine because it has been studied more extensively in pregnancy and can be dosed twice per day more predictably than glargine, as previously mentioned. (See "General principles of insulin therapy in diabetes mellitus", section on 'Safety'.)

Oral hypoglycemic agents — Metformin and glyburide are the only noninsulin antihyperglycemic drugs used in pregnancy. Both oral hypoglycemic agents offer the advantage of significantly decreased cost compared with insulin.

Choosing metformin versus glyburide — Clinically important pregnancy outcomes are generally similar for metformin and glyburide, with only limited evidence of benefit of one oral agent over the other.

●Pregnant outcome – When compared with glyburide in meta-analyses of randomized trials, metformin resulted in:

•Similar rates of perinatal mortality, neonatal hypoglycemia [71], and hypertensive disorders of pregnancy [83].

•Lower mean birth weight (mean difference -191 g, 95% CI -288 to -95 g; mean birth weight 3103 to 3360 g versus 3329 to 3463 g with glyburide) [69].

•Less macrosomia (OR 0.32, 95% CI 0.08-1.19) and large for gestational age infants (OR 0.38, 95% CI 0.18-0.78) [69].

•Less gestational weight gain (mean difference -2.22 kg, 95% CI -3.88 to -0.56 kg) [69].

●Need for supplemental insulin – The frequency of treatment failure (inability to maintain glucose levels in the target range) is similar for glyburide and metformin and ranges from approximately 15 to 30 percent in most trials directly comparing the two drugs [71,84].

●Placental transfer – Both drugs cross the placenta (in contrast to insulin). Fetal metformin levels are 200 percent of the maternal level and glyburide levels are 70 percent of the maternal level, which has unknown long-term consequences [85-87]. Although metformin and glyburide have not been associated with an increased risk of congenital anatomic anomalies, when either drug is prescribed, patients should be made aware that information regarding the long-term effects of transplacental passage, including possible fetal programming effects, are largely unknown, so caution is warranted until more data are available [88-93].

Metformin — A typical dosing regimen is to start metformin extended release (XR) 500 mg orally once daily (with dinner) and, if tolerated, increase by 500 mg (eg, 1000 mg with dinner or 500 mg with dinner plus 500 mg with breakfast) based on the degree of glucose elevations. The dose can then be increased by 500 to 1000 mg orally per week until reaching the usual effective dose of 1500 to 2000 mg orally per day divided into two doses (maximum daily dose is 2500 mg) [94]. An immediate release preparation is also available, but we prefer the XR as it may cause fewer gastrointestinal side effects and fewer daily doses may be needed.

The most common side effects of metformin are gastrointestinal, including a metallic taste in the mouth, mild anorexia, nausea, abdominal discomfort, and soft bowel movements or diarrhea. These symptoms are usually mild, transient, and reversible after dose reduction or discontinuation of the drug. Symptoms can be mitigated by starting at a low dose with slow-dose escalation as needed. In a clinical trial, only 2 percent of study subjects discontinued metformin because of the gastrointestinal side effects [94].

The ADA recommends avoiding metformin in patients with hypertension, preeclampsia, or at risk for intrauterine growth restriction because metformin suppresses mitochondrial respiration, which may adversely affect function, growth, or differentiation of fetal or placental tissues [23,88]; however, any clinical impact of this effect has not been observed in human pregnancies. The American College of Obstetricians and Gynecologists (ACOG) and the Society for Maternal-Fetal Medicine do not include this caveat in their recommendations.

Glyburide — Starting doses of 2.5 to 5 mg once daily are commonly used, increased as needed to a maximum of 20 mg per day. Twice-daily dosing is often necessary to maintain glucose levels in the target range. One group that investigated glyburide pharmacokinetics in pregnancy suggested pregnant patients take the drug 30 to 60 minutes before a meal, rather than with the meal, to improve efficacy [95]. In this study, plasma glyburide concentrations in pregnant patients with GDM did not increase until one hour after drug ingestion, peaked at two to three hours, and returned to baseline by 8 to 10 hours. Thus, the drug took longer to reach peak concentration and was metabolized more rapidly than in nonpregnant females.

Maternal hypoglycemia is the most common side effect, and the risk was higher than that in patients with GDM using insulin in a large trial (28.8 versus 3.5 percent; difference 25.3 percent, 95% CI 16.6-34.0) [73].

Patients who fail to achieve glycemic control with oral pharmacotherapy — If oral pharmacotherapy alone does not adequately manage glucose levels, supplemental insulin can be prescribed and may be easier for the patient than switching to a multidose insulin only regimen. In contrast to nonpregnant patients, dual use of oral agents (eg, metformin plus glyburide) is not recommended in pregnancy because of minimal safety and efficacy data [84] and concerns about adverse fetal effects since both drugs cross the placenta.

OBSTETRIC MANAGEMENT — Obstetric management of the pregnancy is discussed separately. (See "Gestational diabetes mellitus: Obstetric issues and management".)

INTRAPARTUM MANAGEMENT — Intrapartum glucose and insulin management are discussed in detail separately. (See "Pregestational (preexisting) and gestational diabetes: Intrapartum and postpartum glycemic control".)

MATERNAL PROGNOSIS — Most patients with GDM are normoglycemic after delivery. However, they are at high risk for recurrent GDM, prediabetes (impaired glucose tolerance or impaired fasting glucose), and overt diabetes over the subsequent five years. Optimum interpregnancy care to minimize these risks has not been well studied in randomized trials [96]. Feasibility trials of a web-based lifestyle intervention and a telephone-based intervention reported less postpartum weight retention in patients with GDM assigned to the intervention, suggesting this type of behavioral intervention may have a favorable impact [97,98].

Recurrence — GDM in one pregnancy is a strong predictor of recurrence in a subsequent pregnancy [99]. In a study including over 65,000 pregnancies, the frequency of GDM in the second pregnancy among patients with and without previous GDM was 41 and 4 percent, respectively [100]. Risk factors for recurrence include high birth weight in the index pregnancy, older maternal age, high parity, high prepregnancy weight, and high weight between pregnancies [101,102].

Long-term risk — A history of GDM is predictive of an increased risk of developing type 2 diabetes, metabolic syndrome, cardiovascular disease (CVD), and even type 1 diabetes. These risks appear to be particularly high in patients with both GDM and a hypertensive disorder of pregnancy [103]. GDM has been called a "marker," "stress test," or "window" for future diabetes and CVD; it is not considered causal.

●Impaired glucose tolerance – As many as 30 percent of patients with GDM have impaired glucose tolerance during the early postpartum period [104-106].

●Metabolic syndrome – At ≥3 months postpartum, patients with GDM are more likely to have metabolic syndrome, an atherogenic lipid profile, and early vascular dysfunction than those without previous GDM [107-110]. In one study of patients with mild GDM (ie, normal fasting glucose level on glucose tolerance test [GTT]), approximately one-third developed metabolic syndrome within 5 to 10 years of delivery [110].

●Type 2 diabetes – In a meta-analysis, patients with GDM were at an almost 10-fold higher risk of developing subsequent type 2 diabetes than patients with normoglycemic pregnancies (relative risk [RR] 9.51, 95% CI 7.14-12.67; 20 studies including nearly 68,000 patients with GDM and over 1.2 million patients without GDM) [111]. The RR was 17 within the first five years after delivery and approximately 10 after that. Absolute risks for type 2 diabetes at 1 to 5 years, >5 to 10 years, and >10 years postdelivery follow-up were 9, 12, and 16 percent, respectively, compared with 1 to 2 percent in the control groups.

Waist circumference and body mass index (BMI) are the strongest anthropometric measures associated with development of type 2 diabetes in patients with GDM [59,112], as they are in those without GDM. Type 2 diabetes develops in 50 to 75 percent of patients with obesity (BMI ≥30 kg/m²) and a history of GDM versus fewer than 25 percent of those with GDM who achieve normal BMI after delivery [113-115].

Other major risk factors are gestational requirement for insulin and early gestational age at the time of diagnosis (ie, less than 24 weeks of gestation) [112]. Additional risk factors for impaired glucose tolerance and overt diabetes later in life include autoantibodies (eg, glutamic acid decarboxylase, insulinoma antigen-2), high-fasting blood glucose concentrations during pregnancy and early postpartum, higher-fasting plasma glucose at diagnosis of GDM and high glucose levels in the GTT, the number of abnormal values on the glucose tolerance test, neonatal hypoglycemia, and GDM in more than one pregnancy [59,104,105,113,116-120]. In one study, an additional pregnancy increased the rate ratio of type 2 diabetes threefold compared with individuals without an additional pregnancy (RR 3.34, 95% CI 1.80-6.19) [121]. The authors hypothesized that episodes of insulin resistance contribute to the decline in beta-cell function that leads to type 2 diabetes in many high-risk individuals.

Parity, large birth weight, and diabetes in a first-degree relative are less correlated with later diabetes.

●Type 1 diabetes – GDM also appears to be a risk factor for the development of type 1 diabetes, particularly in populations with a high prevalence of this disorder. Specific human leukocyte antigen (HLA) alleles (DR3 or DR4) may predispose to the development of type 1 diabetes postpartum, as does the presence of islet-cell autoantibodies [122-124] or antibodies against glutamic acid decarboxylase or insulinoma antigen 2. GDM in lean pregnant people, need for insulin treatment of GDM, diabetic ketoacidosis during pregnancy, and postpartum hyperglycemia also suggest preexisting unrecognized type 1 diabetes or high risk of developing type 1 diabetes [124]. Although testing for antibodies is not routinely recommended, it is important for clinicians to be aware of this association.

●Cardiovascular disease – Patients with GDM are at higher risk of developing CVD and developing it at a younger age than those with no history of GDM [125-128]. Even mild glucose impairment (defined as an abnormal 50 g one-hour GTT followed by a normal 100 g three-hour GTT) appears to identify individuals at increased risk of future development of CVD, usually myocardial infarction or stroke [129]. Much but not all of this excess risk is related to development of type 2 diabetes later in life. In a pooled analysis of nine studies including over 5 million females and >101,000 cardiovascular events, those with GDM had a twofold higher risk of future CVD compared with those with no history of GDM (RR 1.98, 95% CI 1.57-2.50) [130]. Meta-regression analysis showed that the rates of incident type 2 diabetes across the studies did not affect this risk and when individuals with type 2 diabetes were excluded, GDM was still associated with an increased risk of future CVD (RR 1.56, 95% CI 1.04, 2.32).

Follow-up

Testing — Long-term follow-up for development of type 2 diabetes is routinely recommended for individuals with GDM [23,41]. A common approach is to perform a GTT 4 to 12 weeks after delivery, using the 75 g GTT, as recommended by the American Diabetes Association (ADA) [23]. Criteria for diagnosis of diabetes and prediabetes are shown in the tables (table 2A-B).

Since many obstetrical providers see their patients at four to six weeks postpartum, it makes sense to order the test prior to this visit so the results are available for counseling or to provide an opportunity for scheduling/rescheduling if the test was not performed. Compliance with the 4- to 12-week GTT is poor; however, there is increasing evidence that ordering the test when patients are still hospitalized after birth increases compliance to nearly 100 percent and provides reliable results [106,131]. In an analysis of over 200 patients with GDM who completed a postpartum day 2 75 g GTT, returned for a GTT at postpartum week 4 to 12, and had a A1C checked approximately one year after delivery, there were no significant differences between the day 2 and the 4- to 12-week postpartum GTTs in predicting impaired glucose metabolism (A1C ≥5.7 and <6.5 percent) or diabetes (A1C ≥6.5 percent) at one year [131]. At one year postpartum, the A1C was consistent with impaired glucose metabolism in 35 percent and diabetes in 4 percent of individuals tested.

A fasting plasma glucose level is a reasonable alternative to the GTT but does not allow for diagnosis of impaired glucose tolerance. A glycated hemoglobin (A1C) can be performed in patients in whom obtaining a fasting specimen is especially inconvenient but performs less well for diagnosis of diabetes or prediabetes in postpartum patients because of increased peripartum red cell turnover [132]. (See "Clinical presentation, diagnosis, and initial evaluation of diabetes mellitus in adults", section on 'Diagnostic tests'.)

Breastfeeding during a GTT appears to have a modest lowering effect on glucose levels [133], which could affect interpretation of a borderline test. Patients should be informed in advance that they might need to repeat the test if this happens so they can make an informed decision about breastfeeding during the test versus planning the test at a time/later date when breastfeeding can be avoided. Clinicians should consider retesting those patients whose results fall within 6 mg/dL of the diagnostic cut points.

Counseling

●Patients with prediabetes – Patients who meet criteria for prediabetes are counseled about their subsequent risk for developing overt diabetes and referred for discussion of management options (eg, lifestyle modification such as medical nutritional therapy, use of metformin). They should try to achieve a BMI in the normal range through diet and exercise, and if possible, they should avoid drugs that may adversely affect glucose tolerance (eg, glucocorticoids). They should have yearly assessment of glycemic status. (See "Clinical presentation, diagnosis, and initial evaluation of diabetes mellitus in adults", section on 'Prediabetes' and "Prevention of type 2 diabetes mellitus".)

Breastfeeding may decrease the long-term risk of developing type 2 diabetes. (See "Gestational diabetes mellitus: Obstetric issues and management", section on 'Breastfeeding'.)

●Patients with diabetes – Patients with overt diabetes mellitus should receive appropriate education and treatment. They should also be given advice regarding contraception and the planning of future pregnancies, especially the importance of good glycemic management prior to conception. (See "Overview of general medical care in nonpregnant adults with diabetes mellitus" and "Pregestational (preexisting) diabetes: Preconception counseling, evaluation, and management".)

●Patients with normal test results

•Risk for future diabetes – Patients with a normal GTT should be counseled regarding their risk of developing GDM in subsequent pregnancies and their future risk of developing type 2 diabetes. (See 'Recurrence' above and 'Long-term risk' above.)

•Prevention of future diabetes – Patients should be informed that breastfeeding may decrease their long-term risk of developing type 2 diabetes. (See "Gestational diabetes mellitus: Obstetric issues and management", section on 'Breastfeeding'.)

Lifestyle interventions (diet and exercise) are beneficial for reducing the incidence of type 2 diabetes in persons with prediabetes [134] and these interventions also appear to be beneficial in patients with a history of GDM, whether or not they meet criteria for prediabetes. The annual incidence of diabetes may be reduced by 30 to 50 percent compared with no intervention [135,136].

Drug therapy (eg, metformin, pioglitazone) may also have a role in preventing future type 2 diabetes. In a multicenter randomized trial, both intensive lifestyle and metformin therapy reduced the incidence of future diabetes by approximately 50 percent compared with placebo in patients with a history of GDM; metformin was much more effective than lifestyle intervention in parous patients with previous GDM [135]. This topic is discussed in detail separately. (See "Prevention of type 2 diabetes mellitus".)

•Follow-up laboratory testing – Long-term follow-up is essential. Reassessment of glycemic status should be undertaken at a minimum of every three years [23]. More frequent assessment may be important in patients who may become pregnant again, since early detection of diabetes is important to preconception and early prenatal care. More frequent screening (every one or two years) may also be indicated in patients with other risk factors for diabetes, such as family history of diabetes, obesity, and need for pharmacotherapy during pregnancy.

The best means of follow-up testing has not been defined. The 75 g oral GTT is the more sensitive test for diagnosis of diabetes and impaired glucose tolerance in most populations, but the fasting plasma glucose is more convenient, specific, and reproducible, and less expensive. A1C is convenient and the preferred test for patients who have not fasted overnight. (See "Screening for type 2 diabetes mellitus", section on 'Screening tests'.)

•Prevention of future cardiovascular disease – Given increasing evidence of an association between GDM and future CVD [137], even in the absence of progression to type 2 diabetes, it is reasonable to discuss healthy lifestyle behaviors (eg, heart-healthy diet, maintenance of a healthy weight, tobacco avoidance, and physical activity) with all patients who have had GDM [138]. (See "Overview of primary prevention of cardiovascular disease".)

Follow-up of patients not screened for GDM — In patients who did not undergo screening for GDM, but diabetes is suspected postpartum because of newborn outcome (eg, hypoglycemia, macrosomia, congenital anomalies), a postpartum GTT may be considered. A normal postpartum GTT excludes the presence of type 1 or type 2 diabetes or prediabetes; it does not exclude the possibility of GDM during pregnancy and the future risks associated with this diagnosis. Indications for screening and tests used for screening are discussed separately. (See "Screening for type 2 diabetes mellitus".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Diabetes mellitus in pregnancy".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topics (see "Patient education: Gestational diabetes (diabetes that starts during pregnancy) (The Basics)")

●Beyond the Basics topics (see "Patient education: Gestational diabetes (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●Glucose monitoring – Glucose levels are monitored several times daily in patients with gestational diabetes mellitus (GDM). We suggest self-monitoring of blood glucose levels before breakfast and at one or at two hours after the beginning of each meal. (See 'Glucose monitoring' above.)

●The frequency of testing may be decreased to every other day in specific patients with mild GDM (defined as no more than intermittent glucose elevations within 5 to 10 points above targets), no signs of fetal overgrowth (defined as abdominal circumference (AC) >75^th percentile or estimated fetal weight (EFW) ≥90^th percentile), and normal amniotic fluid volume (ie, no polyhydramnios). (See 'Can the frequency of self-monitoring be reduced?' above.)

●Glucose targets (see 'Glucose target' above)

•Fasting blood glucose concentration: <95 mg/dL (5.3 mmol/L)

•One-hour postprandial blood glucose concentration: <140 mg/dL (7.8 mmol/L)

•Two-hour postprandial glucose concentration: <120 mg/dL (6.7 mmol/L)

●Treatment

•A program of medical nutritional therapy, self-monitoring of blood glucose levels, and pharmacotherapy, when needed, improves some perinatal outcomes (reduction in preeclampsia, macrosomia, and shoulder dystocia). Moderate exercise also improves glycemic control and should be part of the treatment plan for patients with no medical or obstetric contraindications to this level of physical activity. (See 'Rationale for treatment' above and 'Exercise' above.)

•Medical nutritional therapy is the initial approach. Calories are generally divided over three meals and two to four snacks per day and are composed of approximately 40 percent carbohydrate, 20 percent protein, and 40 percent fat. (See 'Medical nutritional therapy' above.)

•Pharmacotherapy is prescribed for patients who do not achieve adequate glycemic control with nutritional therapy and exercise alone (ie, at least 30 percent of glucose levels meet or exceed target thresholds within any one week) (algorithm 1). We also suggest pharmacotherapy for patients with indirect evidence of fetal hyperinsulinemia (eg, AC >75^th percentile or EFW ≥90^th percentile on an early third-trimester sonogram) regardless of maternal glucose levels (Grade 2C). Pharmacotherapy can reduce the occurrence of macrosomia and large for gestational age in newborns. (See 'Indications for pharmacotherapy' above.)

•We suggest prescribing insulin rather than noninsulin antihyperglycemic drugs during pregnancy (Grade 2C). We start with the simplest insulin regimen likely to be effective based on the glucose levels recorded in the patient's blood glucose log and increase the complexity as needed. An alternative approach based on both patient weight and glucose levels is somewhat more complex and likely most appropriate for individuals whose glucose levels are not well managed with simpler paradigms. (See 'Insulin' above.)

•Metformin and glyburide are the only oral antihyperglycemic drugs used in pregnancy and either is a reasonable alternative for patients who decline to take, or are unable to comply with, insulin therapy. The long-term effects of transplacental passage of noninsulin antihyperglycemic agents are not known. (See 'Oral hypoglycemic agents' above.)

●Prognosis – Most patients with gestational diabetes mellitus are normoglycemic after delivery but are at high risk for developing recurrent gestational diabetes mellitus, prediabetes (impaired glucose tolerance or impaired fasting glucose), and overt diabetes. (See 'Recurrence' above and 'Long-term risk' above.)

●Postpartum testing – Patients with GDM should be tested for type 2 diabetes mellitus after pregnancy (table 2A-B). Screening is performed at 4 to 12 weeks postpartum and, if negative, at least every three years thereafter. Lifestyle interventions (eg, achieving a healthy weight, appropriate level of physical activity/exercise) are beneficial for reducing the incidence of type 2 diabetes and related comorbidities such as cardiovascular disease (CVD). (See 'Maternal prognosis' above.)

ACKNOWLEDGMENTS — The UpToDate editorial staff acknowledges Lois Jovanovic, MD, Donald R Coustan, MD, and Michael Greene, MD, who contributed to earlier versions of this topic review.