| | |

Received: 21 December 2016    Revised: 14 March 2017    Accepted: 8 May 2017    First published online: 13 June 2017

DOI: 10.1002/ijgo.12205

CLINICAL ARTICLE
Obstetrics

Prevalence of gestational diabetes mellitus evaluated by

universal screening with a 75-­g, 2-­hour oral glucose tolerance
test and IADPSG criteria

Deniz Karcaaltincaba1 | Pinar Calis1,* | Nagehan Ocal2 | Aykut Ozek1 | 

Melis Altug Inan1 | Merih Bayram1

1
Department of Obstetrics and
Gynecology, Gazi University Faculty of Abstract
Medicine, Ankara, Turkey Objective: To determine the prevalence of gestational diabetes mellitus (GDM) and its
2
Gazi University Faculty of Medicine, Ankara,
association with maternal age among Turkish women diagnosed by International
Turkey
Association of Diabetes and Pregnancy Study Group (IADPSG) criteria.
*Correspondence
Methods: A cross-­sectional study was conducted in 2013–2015 among non-­diabetic preg-
Pinar Calis, Department of Obstetrics and
Gynecology, Gazi University Faculty of nant women aged 18–49 years who were universally screened for GDM by IADPSG crite-
Medicine, Besevler, Ankara, Turkey.
ria. The percentage of women meeting each diagnostic threshold and the prevalence of
Email: [email protected]
GDM by age group were calculated. Linear trends were evaluated by logistic regression.
Results: Among 1434 women screened, 159 (11.1%, 95% confidence interval 9.5%–
12.7%) were diagnosed with GDM; eleven of these women had been diagnosed accord-
ing to a fasting glucose level in the first trimester. The prevalence of GDM was 6.6%
(10/151), 7.3% (37/507), 8.8% (42/479), 16.7% (45/270), and 35.2% (25/71) among
women aged younger than 25, 25–29, 30–34, 35–39, and 40 years or older, respectively.
GDM prevalence increased with age (P<0.001). The numbers of women diagnosed with
GDM in the second trimester who exceeded one, two, and three thresholds of the 2-­hour
oral glucose tolerance test were 66 (44.6%), 52 (35.1%), and 30 (20.3%), respectively.
Conclusion: Prevalence of GDM was correlated with maternal age. Most women diag-
nosed in the second trimester exceeded the threshold at only one of the three
timepoints.

KEYWORDS
Gestational diabetes mellitus; IADPSG criteria; International Association of Diabetes and
Pregnancy Study Group; Maternal age; Pregnancy

1 | INTRODUCTION to hyperglycemia and high insulin levels, which might be related to

increased risk of childhood and adult obesity, and of GDM and type 2
Gestational diabetes mellitus (GDM) is carbohydrate intolerance of vari- diabetes at childbearing ages.4 Therefore, diagnosis of GDM is crucial.
1
able severity that has its onset or is recognized during pregnancy. It is GDM remains one of the most controversial syndromes in obstet-
1
one of the most common medical disorders of pregnancy. Women with rics because of the different screening protocols, diagnostic tests,
GDM are at high risk for future type 2 diabetes, with 50% of women and diagnostic criteria applied worldwide. At present, the American
with GDM subsequently developing type 2 diabetes within 25 years.2,3 College of Obstetricians and Gynecologists (ACOG) recommends
Additionally, maternal diabetes during pregnancy exposes the fetus selective screening with a two-­step approach between weeks 24 and

148  |  wileyonlinelibrary.com/journal/ijgo

© 2017 International Federation of Int J Gynecol Obstet 2017; 138: 148–151
Gynecology and Obstetrics
Karcaaltincaba ET AL |
      149

28 of pregnancy.1 The first step involves the determination of venous The mean prevalence of GDM and the percentage of women that
glucose levels 1 hour after oral ingestion of 50 g of glucose solution. met or exceeded each of the 75-­g OGTT thresholds were calculated.
Individuals exceeding the threshold (7.77 mmol/L [140 mg/dL]) undergo Additionally, the percentages of women meeting one, two, or all three
a 100-­g, 3-­hour diagnostic oral glucose tolerance test (OGTT). In 1979, thresholds were compared.
the National Diabetes Data Group (NDDG) recommended threshold The study women were divided into five subgroups by maternal
5
values for 100-­g OGTT, and these were accepted as diagnostic cutoff age (<25, 25–29, 30–34, 35–39, and ≥40 years), and the prevalence
points until 1999. In 2000, Carpenter and Coustan (CC) threshold val- of GDM in each subgroup was calculated separately. The linear trend
ues6 were adopted worldwide. in the prevalence of GDM with age was evaluated by logistic regres-
In 2010, however, the International Association of Diabetes and sion. Statistical analysis was performed using SPSS version 21.0 (IBM,
Pregnancy Study Group (IADPSG) recommended screening all women Armonk, NY, USA). P<0.05 was considered statistically significant.
using a one-­step, 75-­g, 2-­h diagnostic OGTT and new diagnostic cri-
teria based on data from the Hyperglycemia and Adverse Pregnancy
Outcome (HAPO) study.7–9 The same approach has also been adopted 3 | RESULTS
by the American Diabetes Association. It has been predicted that,
in the US population, the mean prevalence of GDM diagnosed by During the study period, 1478 consecutive pregnant women were enrolled
the proposed IADPSG criteria will be approximately 18%, which is in the study. Eleven of them had a fasting glucose level of 5.11 mmol/L
at least twofold higher than the prevalence determined by ACOG (92 mg/dL) or higher at the first prenatal visit and were diagnosed with
1
recommendations. GDM. The remaining 1467 patients were scheduled to undergo the 75-­g
The aim of the present study was to determine the prevalence of 2-­hour OGTT at 24–28 weeks. Of the 1423 women who completed the
GDM and its association with maternal age in a Turkish population, as 75-­g OGTT test, 148 were diagnosed with GDM. Thus, the overall mean
determined by universal screening using the IADPDG criteria. This will prevalence of GDM was 11.1% (159/1434) (Fig. 1).
demonstrate how these criteria affect both the prevalence of GDM Overall, OGTT thresholds were exceeded 260 times (fasting,
and the proportion of pregnant women needing treatment due to glu- 1-­hour, and 2-­hour values considered separately). The most common
cose intolerance. point at which glucose values were elevated was at 1 hour (108/260
[41.5%]) (Fig. 2). The percentage of women who met the GDM diag-
nosis with only one threshold value (fasting, 1-­hour, or 2-­hour) was
2 | MATERIALS AND METHODS 44.6% (66/148), the percentage with two threshold values (fasting
and 1-­hour, fasting and 2-­hour, or 1-­hour and 2-­hour) was 35.1%
The present cross-­sectional study was undertaken at the Department (52/148), and the percentage with all three threshold values was
of Obstetrics and Gynecology, Gazi University Faculty of Medicine, 20.3% (30/148) (Table 1).
Ankara, Turkey. Consecutive non-­diabetic pregnant women aged
18–49 years were enrolled prospectively between January 1, 2013,
Pregnant women screened
and December 31, 2015. Pregnant women with known type 1 or for GDM (n=1478)
type 2 diabetes were excluded. Approval for the study was obtained
from the Institutional Review Board in Clinical Studies (Approval No.
First prenatal visit
14.12.2015-­130). All participants gave informed consent.
Since 2011, the study clinic has followed the recommendations
of the American Diabetes Association and IADPSG for GDM screen-
Fasting plasma glucose ≥5.11 mmol/L Fasting plasma glucose <5.11 mmol/L
ing, including both the first-­trimester fasting glucose level and the but no diagnosis of overt diabetes (n=1467)
(n=11)
one-­step approach using the 75-­g, 2-­hour diagnostic OGTT at weeks Excluded (n=44)
Refused testing (n=23)
24–28. Universal screening for GDM was performed at the first prena- Vomiting (n=21)

tal visit via a fasting glucose test. Women whose fasting glucose levels
75-g OGTT at 24-28 wk
were met or exceeded 5.11 mmol/L (92 mg/dL) were considered to (n=1423)

have GDM. The remaining women received an appointment for the

75-­g, 2-­hour OGTT at 24–28 weeks of pregnancy.
GDM diagnosed (n=148)
To determine fasting glucose levels in the OGTT, venous plasma
samples were collected from the women after a 12-­hour overnight
fast. The women were then given 75 g of anhydrous glucose orally,
GDM by IADPSG criteria
and 1-­hour and 2-­hour venous plasma samples were collected. Blood (n=159, 11.1%)

samples were collected in fluoride-­oxalate vials. GDM was diagnosed

F I G U R E   1   Flow chart showing recruitment of the study women
if at least one of the threshold values was met or exceeded as per the
and prevalence of GDM by IADPSG criteria. Abbreviations: GDM,
IADPSG criteria (fasting value 5.11 mmol/L [92 mg/dL]; 1-­hour value gestational diabetes; IADPSG, International Association of Diabetes
9.99 mmol/L [180 mg/dL]; 2-­hour value 8.49 mmol/L [153 mg/dL]). and Pregnancy Study Group.
 |
150       Karcaaltincaba ET AL

T A B L E   2   Prevalence of GDM diagnosed by IADPSG criteria

stratified by maternal age.

Total no. of No. of women % of women with

Age, y women with GDM GDM (95% CI)

<25 151 10 6.6 (2.6–10.7)

25–29 507 37 7.3 (5.1–9.7)
30–34 479 42 8.8 (6.2–11.3)
35–39 270 45 16.7 (12.2–21.2)
≥40 71 25 35.2 (24.2–47.3)
Total 1478 159 11.1 (9.5–12.7)

Abbreviations: GDM, gestational diabetes; IADPSG, International

F I G U R E   2   Percentage of positive results occurring at each of the
Association of Diabetes and Pregnancy Study Group; CI, confidence
OGTT timepoints. Abbreviation: OGTT, oral glucose tolerance test.
interval.

The prevalence of GDM was correlated with maternal age

Studies have been done in several countries to determine the prev-
(P<0.001), with the highest prevalence among women aged 40 years
alence of GDM in accordance with the new IADPSG criteria. The preva-
or older (Table 2).
lence of GDM diagnosed by IADPSG criteria has been reported as 17.6%,
4.2%, 11.8%, 9.5%, 23.3%, 8.6%, and 45% in Singapore,11 Greenland,12
4 |  DISCUSSION Switzerland,13 South Korea,14 Sri Lanka,15 Sub-­Saharan Africa,16 and the
United Arab Emirates,17 respectively. By contrast, a two-­step approach
Among the study women, the mean prevalence of GDM diagnosed by for diagnosis determined a GDM prevalence of 3.3% in Switzerland13
IADPSG criteria was 11.1%. GDM prevalence increased with age: the and 12% in the United Arab Emirates.17 This approximately fourfold
lowest frequency was observed among women younger than 25 years increase in prevalence of GDM determined by the IADPSG criteria is
(6.6%) and the highest among those aged 40 years or older (35.2%). similar to the present findings. The main criticism of ACOG regarding the
Of the three threshold values in the 75-­g 2-­hour diagnostic OGTT, the IADDPG criteria, as compared with the conventional two-­step screen-
elevated 1-­hour value was found to be the most prevalent (40.1%). ing approach, has been the increased cost of health care owing to the
In a previous study using NDDG and CC criteria,10 the mean higher numbers of women diagnosed with GDM.1
prevalence of GDM in the Turkish population was reported as 3.17% GIGT is diagnosed on the basis of one positive value in a 100-­g
and 4.48%, and the mean prevalence of gestational impaired glucose OGTT. Although not accepted to have GDM, women with GIGT have
tolerance (GIGT) as 1.97% and 2.46%, respectively. The prevalence an increased likelihood of developing abnormal glucose tolerance later
of both GDM and GIGT was correlated with maternal age.9 In the in life, similar to the increased risk for women with GDM.18 Additionally,
present study, on the basis of the one-­step screening approach, the GIGT is associated with increased risk of adverse perinatal outcomes;
mean prevalence of GDM was 11.1% (95% confidence interval 9.5%– thus, dietary counseling and glucose monitoring are recommended to
12.7%), showing an approximately 3.5-­fold increase as compared with reduce perinatal morbidity.19 Given the increased number of women
the NDDG criteria, and a 2.5-­fold increase as compared with the CC needing treatment during pregnancy for GDM identified by the IADPSG
criteria. criteria, cases of GIGT should also be evaluated by these criteria.
In a previous study,10 the percentage of pregnant women needing
T A B L E   1   Women with gestational diabetes meeting or exceeding
treatment due to glucose intolerance (GDM+GIGT) was 5.14% and
each IADPSG threshold (n=148).a
6.94%, respectively. Although the IADPSG criteria led to a 2.5-­fold
No. (%) of women and 3.5-­fold increase in GDM prevalence relative to the NDDG and CC
Glucose test meeting threshold
criteria, respectively, it caused only a 1.6-­fold and 2.1-­fold increase in
Fasting 20 (13.5) the respective frequency of pregnant women needing treatment due
Only 1-­h 33 (22.3) to glucose intolerance. Although the difference in prevalence of GDM
Only 2-­h 13 (8.8) diagnosed by the IADPDG criteria was 7.93% and 6.62% (vs NDDG
Fasting, 1-­h 13 (8.8) and CC criteria, respectively), the difference in pregnant women need-
Fasting, 2-­h 7 (4.7) ing treatment due to glucose intolerance was only 5.96% and 4.16%

1-­h, 2-­h 32 (21.6) (vs NDDG and CC criteria, respectively).

The increase in prevalence of GDM with age observed in the pres-
Fasting, 1-­h, 2-­h 30 (20.3)
ent study is similar to previous reports using different screening crite-
Abbreviation: IADPSG, International Association of Diabetes and
ria.14,20 In a comparison of the prevalence of GDM or GDM plus GIGT
Pregnancy Study Group.
a
Fasting threshold 5.11 mmol/L (92 mg/dL); 1-­h value 9.99 mmol/L across different age groups using the present findings and data obtained
(180 mg/dL); 2-­h value 8.49 mmol/L (153 mg/dL). by NDDG and CC criteria,10 the highest frequencies were observed
Karcaaltincaba ET AL |
      151

for women younger than 25 years of age (6.6%, 0.95%, and 1.44% by 2. England LJ, Dietz PM, Njoroge T, et  al. Preventing type 2 diabetes:
IADPSG, NDDG, and CC criteria, respectively). In a previous comparison Public health implications for women with a history of gestational dia-
betes mellitus. Am J Obstet Gynecol. 2009;200:365.e361–365.e368.
of CC criteria and NDDG criteria,10 the largest difference in prevalence
3. O’Sullivan JB. Body weight and subsequent diabetes mellitus. JAMA.
was also found for women younger than 25 years9; this discrepancy was 1982;248:949–952.
attributed to the increased sensitivity of the test using CC criteria due 4. Dabelea D, Hanson RL, Lindsay RS, et  al. Intrauterine exposure to
to its lower threshold and a markedly lower incidence of GDM at young diabetes conveys risks for type 2 diabetes and obesity: A study of
discordant sibships. Diabetes. 2000;49:2208–2211.
ages. These observations infer that the IADPSG criteria might have even
5. National Diabetes Data Group. Classification and diagnosis of dia-
higher sensitivity than the CC criteria in diagnosing GDM. betes mellitus and other categories of glucose intolerance. Diabetes.
In the present study, GDM was diagnosed when one glucose value 1979;28:1039–1057.
met or exceeded the criteria defined for a positive test during the 6. Gestational diabetes mellitus. Diabetes Care. 2000;23(Suppl.1):
S77–S79.
75-­g OGTT, as compared with two or more glucose positive values
7. Getahun D, Nath C, Ananth CV, Chavez MR, Smulian JC. Gestational
during the 100-­g OGTT in previous screens. The frequency of women diabetes in the United States: Temporal trends 1989 through 2004.
who were diagnosed with GDM by one elevated threshold value, two Am J Obstet Gynecol. 2008;198:525.e521–525.e525.
elevated threshold values, and three elevated threshold values was 8. Metzger BE, Gabbe SG, Persson B, et al. International association of
diabetes and pregnancy study groups recommendations on the diag-
44.6%, 35.1%, and 20.3%, respectively. Thus, almost half the women
nosis and classification of hyperglycemia in pregnancy. Diabetes Care.
who were diagnosed with GDM had only one elevated threshold value.
2010;33:676–682.
The present results provide a good estimate of the higher prev- 9. Metzger BE, Lowe LP, Dyer AR, et  al. Hyperglycemia and adverse
alence of GDM defined by the IADPSG criteria in Turkey, and the pregnancy outcomes. N Engl J Med. 2008;358:1991–2002.
increased number of patients needing treatment to according to age. 10. Karcaaltincaba D, Kandemir O, Yalvac S, Guvendag-Guven S, Haberal
A. Prevalence of gestational diabetes mellitus and gestational impaired
The study focused on determining the prevalence of GDM and did
glucose tolerance in pregnant women evaluated by National Diabetes
not analyze its consequences, which represents a limitation. In the Data Group and Carpenter and Coustan criteria. Int J Gynecol Obstet.
HAPO study, the threshold levels were derived from an odds ratio of 2009;106:246–249.
1.75 for a macrosomic fetus and clinical hyperinsulinemia.9 As stated 11. de Seymour J, Chia A, Colega M, et  al. Maternal dietary patterns
and gestational diabetes mellitus in a multi-­ethnic Asian cohort: The
in the HAPO study,9 use of the IADPSG diagnostic criteria coupled
GUSTO Study. Nutrients. 2016;8:574.
with proper treatment might lead to a lower incidence of macrosomia, 12. Pedersen ML, Olesen J, Jorgensen ME, Damm P. Gestational diabetes
primary cesarean delivery, clinically diagnosed neonatal hypoglycemia, mellitus in Greenland: A national study of prevalence and testing effi-
delivery before 37 weeks, pre-­eclampsia, shoulder dystocia or birth cacy. Int J Circumpolar Health. 2016;75:32167.
13. Huhn EA, Massaro N, Streckeisen S, et al. Fourfold increase in prev-
injury, hyperbilirubinemia, and need for intensive neonatal care.
alence of gestational diabetes mellitus after adoption of the new
In conclusion, the prevalence of GDM was found to be higher when International Association of Diabetes and Pregnancy Study Groups
assessed by a one-­step screening approach with the IADPSG criteria (IADPSG) criteria. J Perinat Med. 2017;45:359–366.
than with the two-­step approach, and was found to be correlated with 14. Koo BK, Lee JH, Kim J, Jang EJ, Lee CH. Prevalence of gestational dia-
betes mellitus in Korea: A National Health Insurance Database Study.
maternal age. The proportion of women needing treatment owing to
PLoS ONE. 2016;11:e0153107.
glucose intolerance was also higher, but the increase was less marked 15. Meththananda Herath HM, Weerarathna TP, Weerasinghe NP. Is risk
as compared with the increase in GDM prevalence. factor-­based screening good enough to detect gestational diabetes
mellitus in high-­risk pregnant women? A Sri Lankan experience Int J
Prev Med. 2016;7:99.
AUT HOR CONTRI B UTI O N S 16. Olagbuji BN, Atiba AS, Olofinbiyi BA, et al. Prevalence of and risk fac-
tors for gestational diabetes using 1999, 2013 WHO and IADPSG
DK designed and planned the study, and wrote the manuscript. PC criteria upon implementation of a universal one-­step screening and
planned the study, analyzed the data, and wrote the manuscript. NO diagnostic strategy in a sub-­Saharan African population. Eur J Obstet
Gynecol Reprod Biol. 2015;189:27–32.
conducted the study and wrote the manuscript. AO planned and con-
17. Agarwal MM, Dhatt GS, Othman Y. Gestational diabetes: Differences
ducted the study, and revised the manuscript. MAI conducted the between the current international diagnostic criteria and implica-
study and wrote the manuscript. MB designed and planned the study, tions of switching to IADPSG. J Diabetes Complications. 2015;29:
and revised the manuscript. 544–549.
18. Corrado F, D’Anna R, Cannata ML, et al. Positive association between
a single abnormal glucose tolerance test value in pregnancy and
subsequent abnormal glucose tolerance. Am J Obstet Gynecol.
CO NFLI CTS OF I NTE RE S T
2007;196:339.e331–339.e335.
The authors have no conflicts of interest. 19. McLaughlin GB, Cheng YW, Caughey AB. Women with one ele-
vated 3-­hour glucose tolerance test value: Are they at risk for
adverse perinatal outcomes? Am J Obstet Gynecol. 2006;194:
REFERENCES e16–e19.
20. Tamayo T, Tamayo M, Rathmann W, Potthoff P. Prevalence of gesta-
1. American College of Obstetrics and Gynecology. Practice Bulletin No. tional diabetes and risk of complications before and after initiation of
137: Gestational diabetes mellitus. Obstet Gynecol. 2013;122(2 Pt a general systematic two-­step screening strategy in Germany (2012-­
1):406–416. 2014). Diabetes Res Clin Pract. 2016;115:1–8.