The American College of

Obstetricians and Gynecologists

WOMEN’S HEALTH CARE PHYSICIANS

P RACTICE BULLET IN
CLINICAL MANAGEMENT GUIDELINES FOR OBSTETRICIAN – GYNECOLOGISTS

NUMBER 137, AUGUST 2013 (Replaces Practice Bulletin Number 30, September 2001,
Committee Opinion Number 435, June 2009, and
Committee Opinion Number 504, September 2011)

Gestational Diabetes Mellitus

Gestational diabetes mellitus (GDM) is one of the most common medical complications of pregnancy. Debate con-
tinues to surround both the diagnosis and treatment of GDM despite several recent large-scale studies addressing
these issues. The purpose of this document is to 1) provide a brief overview of the understanding of GDM, 2) provide
management guidelines that have been validated by appropriately conducted clinical research, and 3) identify gaps
in current knowledge toward which future research can be directed.

Background its associated potential morbidities. Most importantly,

women with GDM have an increased risk of develop-
Definition and Prevalence ing diabetes later in life. It is projected that up to 50%
of women with GDM will develop diabetes 22–28 years
Gestational diabetes mellitus (GDM) is a condition in after pregnancy (3, 4). The progression to type 2 diabetes
women who have carbohydrate intolerance with onset or may be influenced by ethnicity and the incidence of obe-
recognition during pregnancy. The prevalence of GDM sity. For example, 60% of Latin-American women with
varies in direct proportion to the prevalence of type 2 dia- GDM may develop type 2 diabetes by 5 years after the
betes in a given population or ethnic group. It has been index pregnancy (5).
estimated that up to 6–7% of pregnancies are compli- The offspring of women with GDM are at increased
cated by diabetes mellitus (DM) and that approximately risk of macrosomia, neonatal hypoglycemia, hyperbiliru-
90% of these cases represent women with GDM (1). An binemia, operative delivery, shoulder dystocia, and birth
increased prevalence of GDM is found among Hispanic, trauma. The relationship between maternal hypergly-
African American, Native American, Asian, and Pacific cemia and fetal macrosomia, as well as other adverse
Islander women. With the increase in obesity and seden- outcomes, has been confirmed in the Hyperglycemia and
tary lifestyle, the prevalence of DM among reproductive- Adverse Pregnancy Outcome study (6). This multicenter
aged women is increasing globally. international study demonstrated a continuous relation-
ship between maternal glucose levels and cesarean deliv-
Maternal and Fetal Complications ery, birth weight greater than the 90th percentile, clinical
Women with GDM are at higher risk of gestational hyper- neonatal hypoglycemia, and fetal hyperinsulinemia. An
tension, preeclampsia (2), and cesarean delivery and increase in each of the three values on the 75-g, 2-hour

Committee on Practice Bulletins—Obstetrics. This Practice Bulletin was developed by the Committee on Practice Bulletins—Obstetrics with the
assistance of Mark B. Landon, MD, and Wanda K. Nicholson, MD. The information is designed to aid practitioners in making decisions about appropriate
obstetric and gynecologic care. These guidelines should not be construed as dictating an exclusive course of treatment or procedure. Variations in practice
may be warranted based on the needs of the individual patient, resources, and limitations unique to the institution or type of practice.

406 VOL. 122, NO. 2, PART 1, AUGUST 2013 OBSTETRICS & GYNECOLOGY
oral glucose tolerance test (OGTT) used for GDM diag-
nosis in this study was associated with a graded increase Box 1. Early Screening Strategy for
in these outcomes. Detecting Gestational Diabetes
s 7OMENWITHTHEFOLLOWINGRISKFACTORSARECANDI-
Screening Practices, Diagnostic DATESFOREARLYSCREENING
Thresholds, and Treatment Benefits ˆ0REVIOUSMEDICALHISTORYOFGESTATIONALDIABETES
Historically, screening for GDM consisted of obtaining MELLITUS
the patient’s medical history, relying primarily on past ˆ+NOWNIMPAIREDGLUCOSEMETABOLISM
obstetric outcomes and a family medical history of type
ˆ/BESITYBODYMASSINDEXGREATERTHANOREQUALTO
2 diabetes. In 1973, O’Sullivan and Mahan proposed ;CALCULATEDASWEIGHTINKILOGRAMSDIVIDEDBY
the 50-g, 1-hour oral glucose tolerance test (7). This test HEIGHTINMETERSSQUARED=
has become widely used—an estimated 95% of obstetric
s )FGESTATIONALDIABETESMELLITUSISNOTDIAGNOSED
groups in the United States report performing universal BLOODGLUCOSETESTINGSHOULDBEREPEATEDAT
screening using the 50-g, 1-hour oral glucose tolerance nWEEKSOFGESTATION
test. However, consistent data that demonstrate an over-
all benefit to screening all pregnant women for GDM $ATAFROM-ETZGER"%"UCHANAN4!#OUSTAN$2DE,EIVA!$UNGER$"
are lacking (8). (ADDEN$2ETAL3UMMARYANDRECOMMENDATIONSOFTHE&IFTH
)NTERNATIONAL7ORKSHOP #ONFERENCEON'ESTATIONAL$IABETES-ELLITUS
The use of traditional historic factors (family or ;PUBLISHEDERRATUMAPPEARSIN$IABETES#ARE=$IABETES
personal history of diabetes, previous adverse pregnancy #ARESUPPL 3n,EVEL)))
outcome, glycosuria, and obesity) to identify GDM will
miss approximately one half of women with GDM (9).
It was recognized at the Fifth International Workshop A one-step approach to establishing the diagnosis of
Conference on Gestational Diabetes Mellitus that certain GDM using a 75-g, 2-hour OGTT has been used and pro-
features place women at low risk of GDM, and it may moted by other organizations. In 2010, the International
not be cost-effective to screen this group of women. Association of Diabetes and Pregnancy Study Group
However, such low-risk women represent only 10% convened a workshop conference to recommend new
of the population and selecting these individuals who diagnostic criteria based on the Hyperglycemia and
should not be screened may add unnecessary complexity Adverse Pregnancy Outcome study data (12). Based on
to the screening process (10). expert consensus, an odds ratio of 1.75 (compared with
the population mean) for various adverse outcomes was
used to define blood glucose thresholds for diagnosis of
Clinical Considerations and GDM. The International Association of Diabetes and
Recommendations Pregnancy Study Group recommended that a universal
75-g, 2-hour OGTT be performed during pregnancy
How is gestational diabetes mellitus diagnosed? and that the diagnosis of GDM be established when
any single threshold value on the 75-g, 2-hour OGTT
All pregnant patients should be screened for GDM, was met or exceeded (fasting value, 92 mg/dL; 1-hour
whether by the patient’s medical history, clinical risk value, 180 mg/dL; and 2-hour value, 153 mg/dL) (12).
factors, or laboratory screening test results to determine Overall, using the proposed International Association
blood glucose levels. Screening is generally performed of Diabetes and Pregnancy Study Group criteria would
at 24–28 weeks of gestation. Early pregnancy screen- identify approximately 18% of the U.S. population as
ing for undiagnosed type 2 diabetes, also is suggested having GDM, although in some subpopulations, the pro-
in women with risk factors, including those with a prior portion of women in whom GDM is diagnosed would be
history of GDM (see Box 1) (11). If the result of early even higher. The American Diabetes Association (ADA)
testing is negative, repeat screening for high-risk women endorsed the International Association of Diabetes and
is recommended at 24–28 weeks of gestation. The two- Pregnancy Study Group criteria while acknowledging
step approach to testing, commonly used in the United that adopting these cutoffs will significantly increase the
States, is based on first screening with the administration prevalence of GDM (11).
of 50 g of an oral glucose solution followed by a 1-hour There are no data from randomized clinical tri-
venous glucose determination. Those individuals meet- als (RCTs) regarding therapeutic interventions for the
ing or exceeding the screening threshold undergo a 100-g, expanded group of women designated as having GDM
3-hour diagnostic OGTT. based on the International Association of Diabetes and

VOL. 122, NO. 2, PART 1, AUGUST 2013 Practice Bulletin Gestational Diabetes Mellitus 407
Pregnancy Study Group criteria. These additional women Table 1. Proposed Diagnostic Criteria for Gestational
in whom GDM would be diagnosed may be at a lower Diabetes Mellitus
risk of adverse outcomes than women in whom GDM
Plasma or Serum Plasma Level
was diagnosed by traditional criteria and may not derive
Glucose Level National Diabetes
similar benefits from interventions (13). Carpenter and Coustan Data Group
In 2013, a Eunice Kennedy Shriver National Insti- Status Conversion Conversion
tute of Child Health and Human Development Consen- mg/dL mmol/L mg/dL mmol/L
sus Development Conference on diagnosing gestational
diabetes recommended that health care providers con- Fasting 95 5.3 105 5.8
tinue to use a two-step approach to screen for and diag- One hour 180 10.0 190 10.6
nose GDM because no evidence exists that using these Two hours 155 8.6 165 9.2
2-hour OGTT criteria to diagnose GDM would lead to Three hours 140 7.8 145 8.0
clinically significant improvements in maternal or new- Adapted with permission from Expert Committee on the Diagnosis and
born outcomes, but would lead to a significant increase Classification of Diabetes Mellitus. Report of the Expert Committee on the
in health care costs (14, 15). The American College of Diagnosis and Classification of Diabetes Mellitus. Expert Committee on the
Diagnosis and Classification of Diabetes Mellitus. Diabetes Care 2000;23
Obstetricians and Gynecologists supports this recommen- (suppl 1):S4–19.
dation and recommends that before the testing approach
and diagnostic criteria for GDM are changed, implications
other. However, given the benefits of standardization,
of such changes should be studied.
practitioners and institutions should select a single set of
Screening thresholds for the 1-hour glucose chal-
diagnostic criteria, either plasma or serum glucose levels
lenge have varied from 130 mg/dL to 140 mg/dL, with
designated by the Carpenter and Coustan criteria or the
varying sensitivities and specificities reported. There are
plasma levels established by the National Diabetes Data
no randomized trials to support a clear benefit to one
Group, for consistent use within their patient populations.
cutoff compared with others. Data also are insufficient
In one cross-sectional study that compared the two sets
with regard to pregnancy outcomes to determine an ideal
of criteria in more than 26,000 women, the diagnosis of
threshold value, although standardization of a screening
GDM increased, on average, by 50% with the use of the
threshold has been recently recommended (14). A value
Carpenter and Coustan thresholds (20). Considerations
of 140 mg/dL has been shown in one cohort study to have
for selection of one set of diagnostic criteria over the
lower false-positive rates and improved positive predic-
other could include, but are not limited to, the baseline
tive values across various ethnic groups. In this analy-
prevalence of diabetes in their specific communities and
sis, sensitivities were only marginally improved when
the availability of resources to appropriately manage the
using lower thresholds (130 mg/dL and 135 mg/dL)
numbers of women in whom GDM was diagnosed by any
(16). Establishing a higher standardized threshold of
given protocol. This approach, while imperfect, avoids
140 mg/dL might identify those women at greater risk
establishment of a single set of diagnostic criteria across
of adverse pregnancy outcomes; it may also lower the
all populations based on expert opinion alone.
rate of false-positive screening results and unnecessary
administration of 3-h OGTTs, which have been shown
What is the benefit of treatment of gestational
to be associated with increased maternal stress and dis-
satisfaction regarding the process of screening for and
diabetes mellitus?
diagnosing GDM, in general (17–19). However, in the The 2005 Australian Carbohydrate Intolerance Study in
absence of clear evidence supporting a cutoff of 135 mg/dL Pregnant Women trial was the first large-scale (1,000
versus 140 mg/dL for the 1-h glucose screening test, it is women) randomized treatment trial for GDM (21).
suggested that health care providers select one of these Treatment was associated with a significant reduction
as a single consistent cutoff for their practice, with factors in the rate of the primary outcome, a composite of seri-
such as community prevalence rates of GDM considered ous complications (perinatal death, shoulder dystocia,
in that decision. and birth trauma, including fracture or nerve palsy).
Table 1 lists the diagnostic thresholds established Treatment also reduced the frequency of large for ges-
by both the National Diabetes Data Group and those tational age (LGA)-infants from 22% to 13% and of
established by Carpenter and Coustan, with the latter birth weight greater than 4,000 g from 21% to 10%.
using lower thresholds and subsequently resulting in Among maternal outcomes, preeclampsia was signifi-
higher rates of GDM diagnoses. In the absence of clear cantly reduced with treatment (18% versus 12%).
comparative trials, one set of diagnostic criteria for the The Australian Carbohydrate Intolerance Study
3-h OGTT cannot be clearly recommended above the in Pregnant Women was followed by the 2009 report

408 Practice Bulletin Gestational Diabetes Mellitus OBSTETRICS & GYNECOLOGY

of the Eunice Kennedy Shriver National Institute of What nonpharmacologic treatments are
Child Health and Human Development Maternal–Fetal effective in managing gestational diabetes
Medicine Network randomized, multicenter treatment mellitus?
trial of 958 women with mild GDM (22). Although
there were no differences in the frequency of the primary The goal of nutrition therapy in women with GDM is
composite outcome (perinatal death, neonatal hypogly- to achieve normoglycemia, prevent ketosis, provide
cemia, elevated umbilical cord C-peptide level, or birth adequate weight gain, and contribute to fetal well-being.
trauma), several significant differences in secondary The ADA recommends nutritional counseling for all
outcomes were observed with treatment, including a patients with GDM by a registered dietician, if possible,
lower frequency of LGA-infants, lower frequency of with a personalized nutrition plan based on the individ-
birth weight exceeding 4,000 g, and reduced neonatal fat ual’s body mass index. There are some clinical settings
mass. Moreover, cesarean delivery, shoulder dystocia, in which a dietician may not be readily available. In this
and hypertensive disorders were significantly reduced circumstance, the clinician should be able to provide
in women who were treated for GDM. Therefore, based recommendations to the patient by remembering three
on these studies, women in whom GDM is diagnosed major nutritional components: 1) caloric allotment, 2)
should be treated with nutrition therapy and, when nec- carbohydrate intake, and 3) caloric distribution.
essary, medication for both fetal and maternal benefit. A diet composed of 50–60% carbohydrates will
often result in excessive weight gain and postprandial
How should blood glucose be monitored in a hyperglycemia. For this reason, it has been suggested
woman with GDM? that carbohydrate intake be limited to 33–40% of
calories, with the remaining calories divided between
Once a woman with GDM begins nutrition therapy, sur-
protein (20%) and fat (40%) (27). A randomized trial of
veillance of blood glucose levels is required to be certain
99 women with GDM compared a low-glycemic index
that glycemic control has been established. There is
nutrition plan with a conventional high-fiber diet and
insufficient evidence concerning the optimal frequency
found that both produced similar pregnancy outcomes
of blood glucose testing in women with GDM. Based on
(28). Given these findings, as well as the results of
the data available, the general recommendation is four-
other treatment trials, complex carbohydrates may be
times daily glucose monitoring performed as fasting
preferred to simple carbohydrates because they are less
and either 1 hour or 2 hours after each meal. Once the
likely to produce significant postprandial hyperglycemia
patient’s glucose levels are well controlled by her diet,
(29). In practice, three meals and two to three snacks are
the frequency of glucose monitoring can be modified.
recommended to distribute glucose intake and to reduce
Among adults who are not pregnant, diabetes is
postprandial glucose fluctuations.
often managed using preprandial glucose values through-
Although there are multiple RCTs of exercise and
out the day. In pregnancy, elevated postprandial glucose
lifestyle interventions in adults with diabetes who are
levels may be more predictive of the potential for fetal
not pregnant, there are few published exercise trials in
macrosomia and morbidity, compared with fasting or
women with GDM, and most have small sample sizes
preprandial values. Therefore, fasting glucose values
and limited power to show improvement in glucose
alone do not predict the need for pharmacologic therapy.
levels (30–32). In adults with diabetes who are not preg-
In an RCT that compared the value of postprandial and
nant, exercise, particularly weight training, increases
preprandial measurements for blood glucose monitor-
lean muscle mass and improves tissue sensitivity to
ing of women with GDM, use of the 1-h postprandial
insulin. In overweight or obese women with GDM, exer-
measurement for management of GDM was associated
cise also may be able to improve glycemic control and
with better glycemic control, lower incidence of LGA-
facilitate weight loss. Therefore, a moderate exercise
infants, and lower rates of cesarean delivery due to
program as part of the treatment plan for women with
cephalopelvic disproportion (23).
GDM is recommended (11).
Assessment of blood glucose can be undertaken at
either 1 hour or 2 hours postprandially, but no study to What pharmacologic treatments are effective
date has demonstrated the superiority of either approach
in managing gestational diabetes mellitus?
(24–26). Controlled trials to identify ideal glycemic
targets have not been performed. Both the ADA and the When target glucose levels cannot be consistently
American College of Obstetricians and Gynecologists achieved through nutrition and exercise therapy, phar-
recommend a threshold of 140 mg/dL at 1 hour post- macologic treatment is recommended. However, a sys-
prandial or 120 mg/dL at 2 hours postprandial as glyce- tematic review found no conclusive evidence for the
mic targets to reduce the risk of macrosomia (11). threshold value at which clinicians should start medical

VOL. 122, NO. 2, PART 1, AUGUST 2013 Practice Bulletin Gestational Diabetes Mellitus 409
therapy (33). When pharmacologic treatment of GDM cell adenosine triphosphate calcium channel receptors
is indicated, insulin and oral medications are equivalent to increase insulin secretion and insulin sensitivity of
in efficacy, and either can be an appropriate first-line peripheral tissues. It should not be used in patients who
therapy. Insulin has historically been considered the report a sulfa allergy. Metformin is a biguanide that
standard therapy for GDM management in cases refrac- inhibits hepatic gluconeogenesis and glucose absorp-
tory to nutrition therapy. tion and stimulates glucose uptake in peripheral tissues.
Insulin, which does not cross the placenta, can Current evidence from randomized trials and several
achieve tight metabolic control and traditionally has observational studies of oral antidiabetic agents show
been added to nutrition therapy if fasting blood glucose that maternal glucose levels do not differ substantially
levels are persistently greater than 95 mg/dL, if 1-hour between women treated with insulin versus those treated
levels are persistently greater than or equal to 140 mg/dL, with oral agents, and a meta-analysis suggests that there
or if 2-hour levels are persistently greater than or equal to is no consistent evidence of an increase in any acute or
120 mg/dL. These thresholds have been largely extrap- short-term adverse maternal or neonatal outcomes with
olated from recommendations for managing pregnancy the use of glyburide or metformin compared with the use
in women with preexisting diabetes. If insulin is used, of insulin (34). Therefore, both can be considered for
the typical starting total dosage is 0.7–1.0 units/kg daily, glycemic control in women with GDM.
given in divided doses. In cases in which both fasting Three trials that compared glyburide with insulin
and postprandial hyperglycemia are present, a regimen failed to show any significant difference in glycemic
of multiple injections using both intermediate-acting control (35–37). Several observational studies also
insulin and short-acting insulin alone or in combination have reported generally good outcomes with the use
is administered. Regardless of the starting dosage, subse- of glyburide, although 20–40% of women required the
quent dosage adjustments should be based on the blood addition of insulin to maintain good glycemic control
glucose levels at particular times of the day. Insulin (38–42). The usual dosage of glyburide is 2.5–20 mg
analogs, including insulin lispro and insulin aspart, have daily in divided doses, although pharmacokinetic stud-
been used in pregnancy and do not cross the placenta. ies during pregnancy indicate daily doses up to 30 mg
Insulin lispro has a more rapid onset of action than regu- may be necessary to achieve adequate control (43).
lar insulin and may be useful in improving postprandial Metformin is primarily used in women with preges-
glucose concentrations (Table 2). tational diabetes and in women with polycystic ovary
Oral antidiabetic medications (eg, glyburide and syndrome and infertility. For treatment of pregestational
metformin) are being used increasingly in women with diabetes, metformin is often continued during preg-
GDM, although they have not been approved by the nancy and insulin is added as appropriate to the therapy
U.S. Food and Drug Administration for this indication. regimen. In women with polycystic ovary syndrome,
Glyburide is a sulfonylurea that binds to pancreatic beta- metformin is often continued until the end of the first
trimester, with only limited evidence to suggest that such
use decreases the risks of adverse pregnancy outcomes,
including first-trimester loss (44).
Table 2. Action Profile of Commonly Used Insulin Agents
In one large trial, 751 women with GDM were
Peak of Duration of randomly assigned to receive metformin (plus insulin
Onset of Action Action if needed) or insulin therapy. They experienced similar
Type Action (h) (h) rates of a composite outcome of perinatal morbidity, con-
Insulin lispro 1–15 minutes 1–2 4–5 sisting of neonatal hypoglycemia, respiratory distress,
Insulin aspart 1–15 minutes 1–2 4–5 need for phototherapy, birth trauma, prematurity, and
Regular insulin 30–60 minutes 2–4 6–8 low Apgar scores (45). However, one half of the women
Isophane insulin 1–3 hours 5–7 13–18 randomized to receive metformin required insulin sup-
suspension plementation to achieve glycemic control. Another RCT
(NPH insulin) that compared metformin with glyburide for treatment
Insulin zinc of GDM demonstrated that glyburide may be superior
suspension 1–3 hours 4–8 13–20 to metformin in achieving satisfactory glycemic control
Extended insulin (46). In this study, 35% of women randomized to receive
zinc suspension 2–4 hours 8–14 18–30 metformin required insulin therapy compared with 16%
Insulin glargine 1 hour No peak 24 of those who received glyburide.
Modified from Gabbe SG, Graves CR. Management of diabetes mellitus compli- Although concerns have been raised about the
cating pregnancy. Obstet Gynecol 2003;102:857–68. safety of oral antidiabetic agents during pregnancy, one

410 Practice Bulletin Gestational Diabetes Mellitus OBSTETRICS & GYNECOLOGY

RCT that used umbilical cord blood analysis revealed group in which labor was induced at 38–39 weeks of
no detectable glyburide in exposed pregnancies (35). gestation. Although persuasive, these data have not been
However, it has been reported that glyburide does cross confirmed by additional studies. Therefore, in contrast
the placenta (43). Theoretic concerns regarding this to women with well-controlled, pregestational diabe-
issue include whether long-term glucose homeostasis tes, in whom delivery is recommended after 39 weeks
may be affected in exposed offspring. It also is not of gestation and by the estimated date of delivery, no
known whether glyburide can affect the progression to evidence-based recommendation can be made regarding
type 2 diabetes later in life in women who were treated timing of delivery in women with GDM that is con-
during pregnancy. Although current data demonstrate trolled either with a diet and exercise regimen or with
no adverse short-term effects from oral diabetic therapy medication (49).
during pregnancy on maternal or neonatal health, long- There are insufficient data to determine whether
term outcomes have yet to be studied. This might sug- cesarean delivery should be performed in cases of sus-
gest a role for counseling when prescribing oral agents pected macrosomia to reduce the risk of birth trauma.
to women with GDM. Macrosomia is distinctly more common in women with
GDM, and shoulder dystocia is more likely at a given
Is fetal assessment indicated in pregnancies fetal weight in pregnancies complicated by diabetes
complicated by gestational diabetes mellitus? than in pregnancies not complicated by diabetes (50,
51). Therefore, in women with GDM, it is reasonable
Antepartum fetal testing is recommended for patients for obstetricians to assess fetal growth either by ultra-
with pregestational diabetes. Because the increased risk sonography or by clinical examination late in the third
of fetal demise in patients with pregestational diabetes trimester in an attempt to identify macrosomia before
is related to suboptimal glycemic control, it would be delivery. It has been estimated that up to 588 cesarean
expected that women with GDM who have poor glyce- deliveries for an estimated fetal weight of 4,500 g and up
mic control also would be at risk. Therefore, for women to 962 cesarean deliveries for an estimated fetal weight
with GDM with poor glycemic control, fetal surveillance of 4,000 g would be needed to prevent a single case of
may be beneficial. There is no consensus regarding ante- permanent brachial plexus palsy (52). On the basis of
partum testing in women with well-controlled GDM. available data, it is not completely possible to determine
The specific antepartum test and frequency of testing whether the potential benefits of scheduled cesarean
may be chosen according to local practice. delivery at a given estimated fetal weight are similar for
women with GDM and those with preexisting diabetes. It
What are delivery considerations in preg- appears reasonable, therefore, to recommend that women
nancies complicated by gestational diabetes with GDM be counseled regarding the option of sched-
mellitus? uled cesarean delivery when the estimated fetal weight is
4,500 g or more.
Women with GDM with good glycemic control and no
other complications can be managed expectantly. In
How should women with a history of GDM be
most cases, women with good glycemic control who are
screened and counseled postpartum?
receiving medical therapy do not require delivery before
39 weeks of gestation. In a randomized trial in which Although the carbohydrate intolerance of GDM frequently
women with insulin-treated GDM and fetuses believed resolves after delivery, up to one third of affected women
to be of appropriate weight for gestational age were ran- will have diabetes or impaired glucose metabolism at
domized at 38 weeks of gestation to induction of labor postpartum screening, and it has been estimated that
within 1 week or expectant management, there was no 15–50% will develop type 2 diabetes later in life (53–57).
difference in cesarean delivery rates (47). However, the Postpartum screening at 6–12 weeks is recommended for
induction group gave birth to a smaller proportion of all women who had GDM to identify women with DM,
LGA-infants. In a cohort multiple time series study, a impaired fasting glucose levels, or impaired glucose
policy of induction of labor at 38–39 weeks of gestation tolerance (IGT) (Fig. 1) (11). Women with a history
for women with insulin-treated GDM was compared of GDM have a sevenfold increased risk of developing
with the results in expectantly managed historic controls type 2 diabetes compared with women without a his-
(48). There was no significant difference in macroso- tory of GDM (58). Either a fasting plasma glucose test
mia or cesarean delivery rates, but shoulder dystocia or the 75-g, 2-hour OGTT are appropriate for diagnos-
was experienced by 10% of the expectant management ing overt diabetes in the postpartum period. Although
group beyond 40 weeks of gestation versus 1.4% in the the fasting plasma glucose test is easier to perform, it

VOL. 122, NO. 2, PART 1, AUGUST 2013 Practice Bulletin Gestational Diabetes Mellitus 411
 Gestational diabetes

FPG or 75-g, 2-hr OGTT at 6–12 weeks postpartum

Diabetes mellitus Impaired fasting glucose or IGT or both Normal

Refer for diabetes management Consider referral for management Assess glycemic status
Weight loss and physical activity every 3 years
counseling as needed Weight loss and physical activity
counseling as needed
Consider metformin if combined
impaired fasting glucose and IGT
Medical nutrition therapy
Yearly assessment of glycemic status

Fig.1. Management of postpartum screening results.

Abbreviations: FPG, fasting plasma glucose; OGTT, oral glucose tolerance test; IGT, impaired glucose tolerance.
Postpartum screening for abnormal glucose tolerance in women who had gestational diabetes mellitus. ACOG Committee Opinion
No. 435. American College of Obstetricians and Gynecologists. Obstet Gynecol 2009;113:1419–21.

lacks sensitivity for detecting other forms of abnormal

glucose metabolism; results of the OGTT can confirm an
Summary of
impaired fasting glucose level and IGT. Therefore, the Recommendations and
Fifth International Workshop on Gestational Diabetes
Mellitus recommended that women with GDM undergo
Conclusions
a 75-g, 2-hour OGTT at 6–12 weeks postpartum (59). The following recommendation and conclusion are
Women with impaired fasting glucose, IGT, or dia-
based on good and consistent scientific evidence
betes should be referred for therapy. Women with the
(Level A):
former two conditions may respond to lifestyle modi-
fication and pharmacologic interventions to decrease Women in whom GDM is diagnosed should be
incident diabetes. The ADA recommends repeat testing treated with nutrition therapy and, when necessary,
at least every 3 years for women who had a pregnancy medication for both fetal and maternal benefit.
affected by GDM and normal results of postpartum
When pharmacologic treatment of GDM is indicated,
screening (11).
insulin and oral medications are equivalent in efficacy,
For women who may have subsequent pregnancies,
and either can be an appropriate first-line therapy.
screening more frequently has the advantage of detect-
ing abnormal glucose metabolism before pregnancy and
The following recommendations are based on
provides an opportunity to ensure preconception glucose
limited or inconsistent scientific evidence (Level B):
control (59). Women should be encouraged to discuss
their GDM history and need for screening with all of All pregnant patients should be screened for GDM,
their health care providers. whether by the patient’s medical history, clinical

412 Practice Bulletin Gestational Diabetes Mellitus OBSTETRICS & GYNECOLOGY

 risk factors, or laboratory screening test results to
determine blood glucose levels.
Resources
The following resources are for informational purposes only. Referral
Women with GDM should be counseled regarding to these sources and web sites does not imply the endorsement of the
the option of scheduled cesarean delivery when the American College of Obstetricians and Gynecologists. These resouces
are not meant to be comprehensive. The exclusion of a source or web
estimated fetal weight is 4,500 g or more. site does not reflect the quality of that source or web site. Please note
that web sites are subject to change without notice.

The following recommendations and conclusions Perinatology.com. Gestational diabetes: calculation of

are based primarily on consensus and expert caloric requirements and initial insulin dose. Available
opinion (Level C): at: http://www.perinatology.com/calculators/GDM.htm.
Retrieved May 21, 2013. (Level III)
In the absence of clear evidence supporting a cutoff
of 135 mg/dL versus 140 mg/dL for the 1-h glucose National Heart, Lung, and Blood Institute. Calculate
screening test, it is suggested that health care provid- your body mass index. Available at: http://www.nhlbi
ers select one of these as a single consistent cutoff for support.com/bmi. Retrieved May 21, 2013. (Level III)
their practice, with factors such as community preva-
lence rates of GDM considered in that decision.
In the absence of clear comparative trials, one set of
References
diagnostic criteria for the 3-hour OGTT cannot be 1. Wier LM, Witt E, Burgess J, Elixhauser A. Hospitalizations
clearly recommended above the other. However, related to Diabetes in Pregnancy, 2008. HCUP Statistical
given the benefits of standardization, practitioners Brief #102. Rockville (MD): Agency for Healthcare
Research and Quality; 2010. Available at: http://www.
and institutions should select a single set of diagnos- hcup-us.ahrq.gov/reports/statbriefs/sb102.pdf. Retrieved
tic criteria, either plasma or serum glucose levels April 24, 2012. (Level II-3)
designated by the Carpenter and Coustan criteria or 2. Yogev Y, Xenakis EM, Langer O. The association between
the plasma levels established by the National preeclampsia and the severity of gestational diabetes:
Diabetes Data Group, for consistent use within their the impact of glycemic control. Am J Obstet Gynecol
patient populations. 2004;191:1655–60. (Level II-3) [PubMed] [Full Text]
3. England LJ, Dietz PM, Njoroge T, Callaghan WM, Bruce C,
Once a woman with GDM begins nutrition therapy, Buus RM, et al. Preventing type 2 diabetes: public health
surveillance of blood glucose levels is required to be implications for women with a history of gestational dia-
certain that glycemic control has been established. betes mellitus. Am J Obstet Gynecol 2009;200:365.e1–e8.
(Level III) [PubMed] [Full Text]
Women with GDM with good glycemic control and 4. O’Sullivan JB. Body weight and subsequent diabetes mel-
no other complications can be managed expectantly. litus. JAMA 1982;248:949–52. (Level II-2) [PubMed]
In most cases, women with good glycemic control 5. Kjos SL, Peters RK, Xiang A, Henry OA, Montoro M,
who are receiving medical therapy do not require Buchanan TA. Predicting future diabetes in Latino women
delivery before 39 weeks of gestation. with gestational diabetes. Utility of early postpartum glu-
cose tolerance testing. Diabetes 1995;44:586–91. (Level
Postpartum screening at 6–12 weeks is recom- II-3) [PubMed]
mended for all women who had GDM to identify 6. Metzger BE, Lowe LP, Dyer AR, Trimble ER,
women with DM, impaired fasting glucose, or Chaovarindr U, Coustan DR, et al. Hyperglycemia and
IGT. Women with impaired fasting glucose or IGT adverse pregnancy outcomes. HAPO Study Cooperative
or diabetes should be referred for preventive ther- Research Group. N Engl J Med 2008;358:1991–2002.
(Level II-3) [PubMed] [Full Text]
apy. The ADA recommends repeat testing at least
7. O’Sullivan JB, Mahan CM, Charles D, Dandrow RV.
every 3 years for women who had a pregnancy Screening criteria for high-risk gestational diabetic patients.
affected by GDM and normal results of postpartum Am J Obstet Gynecol 1973;116:895–900. (Level II-3)
screening. [PubMed]
8. Screening for gestational diabetes mellitus: U.S.
Preventive Services Task Force recommendation state-
Proposed Performance ment. U.S. Preventive Services Task Force. Ann Intern
Med 2008;148:759–65. (Level III) [PubMed] [Full Text]
Measure
9. Coustan DR, Nelson C, Carpenter MW, Carr SR, Rotondo
Percentage of women in whom GDM is diagnosed who L, Widness JA. Maternal age and screening for gesta-
have postpartum screening for type 2 diabetes tional diabetes: a population-based study. Obstet Gynecol

VOL. 122, NO. 2, PART 1, AUGUST 2013 Practice Bulletin Gestational Diabetes Mellitus 413