Arch Gynecol Obstet
DOI 10.1007/s00404-014-3566-0
GENERAL GYNECOLOGY
Effect comparison of metformin with insulin treatment
for gestational diabetes: a meta-analysis based on RCTs
Genxia Li • Shujun Zhao • Shihong Cui •
Lei Li • Yajuan Xu • Yuanyuan Li
Received: 21 July 2014 / Accepted: 2 December 2014
Ó Springer-Verlag Berlin Heidelberg 2014
Abstract
Purpose To compare the effects of metformin with
insulin on maternal and neonatal outcomes in gestational
diabetes mellitus (GDM).
Methods A literature search in PUBMED, EMBASE,
Science Direct, Springer link, and Cochrane library was
conducted using the following search terms: ‘‘Gestational
Diabetes’’ or ‘‘GDM’’, and ‘‘insulin’’ and ‘‘metformin’’.
Quality assessment of included studies was determined
with Quality Assessment of Diagnostic Accuracy Studies.
Review Manger 5.2 was used to analyze mean difference
(MD)/risk ratio (RR) and 95 % confidence interval (CI) in
random-effects model or fixed-effects model depending on
the level of heterogeneity.
Results A total of 11 studies were identified. There was
no significant difference of the effect on maternal outcomes
between the two treatments in glycohemoglobin A1c levels
(P = 0.37), fasting blood glucose (P = 0.66), and the
incidence of preeclampsia (P = 0.26); whereas, signifi-
cantly reduced results were found in the metformin group
in pregnancy-induced hypertension (PIH) rate (RR = 0.53,
95 % CI 0.31–0.90, P = 0.02), average weight gains after
enrollment (MD = -1.28, 95 % CI -1.54 to -1.01,
P \ 0.0001), and average gestational ages at delivery
(MD = 0.94, 95 % CI -0.21 to -0.01, P = 0.03).
Regarding neonatal outcomes, when compared with insulin
group, metformin presented significantly lower average
birth weights (MD = -44.35, 95 % CI -85.79 to -2.90,
P = 0.04), incidence of hypoglycemia (RR = 0.69, 95 %
G. Li
S. Zhao
S. Cui (&)
L. Li
Y. Xu
Y. Li
Department of Obstetrics and Gynecology, The Third Affiliated
Hospital of Zhengzhou University, 7 Kangfu Road,
Zhengzhou 450052, People’s Republic of China
e-mail: ShihongCuishc@163.com
CI 0.55–0.87, P = 0.001) and neonatal intensive care unit
(NICU) (RR = 0.82, 95 % CI 0.67–0.99, P = 0.04).
Conclusion Metformin can significantly reduce several
adverse maternal and neonatal outcomes including PIH
rate, incidence of hypoglycemia and NICU, thus it may be
an effective and safe alternative or additional treatment to
insulin for GDM women.
Keywords Gestational diabetes
Metformin
Insulin

Meta-analysis
Introduction
Gestational diabetes mellitus (GDM) is defined as any
degree of glucose intolerance with onset or first recognition
during pregnancy [1]. Its incidence has been increasing
considerably in recent years due to the epidemic of obesity
and type 2 diabetes continues [2–4]. It is well documented
that GDM is associated with several adverse maternal and
neonatal outcomes including macrosomia, primary cesar-
ean delivery, neonatal hypoglycemia, preeclampsia, and
shoulder dystocia [5]. To reduce the risk of these compli-
cations, three strategies including dietary modification,
regulated exercise and medication treatment are currently
applied for the treatment of GDM [6–8]. Because the lack
of glucose control in GDM is associated with higher
maternal and neonatal morbidity, it is essential to effec-
tively control the glucose levels. Medication-based treat-
ment was recommended when the other two strategies have
failed to provide adequate glycemic control.
With the advantage of achieving tight maternal glucose
control without the risk of transferring insulin across the
placenta, insulin has been regarded as the standard for GDM
management [9–11]. However, several disadvantages of
123

insulin treatment are recognized including multiple daily
injections, increased risk of maternal weight gain, risk inci-
dence of hypoglycemia, as well as higher cost [12]. More-
over, the dose of insulin requires modification based on the
patient’s body mass index (BMI), glucose levels and life-
style, which requires detailed guidance for dose change to
ensure safe administration of insulin [13]. However, it was
suggested that the use of safe and effective oral agents may
offer advantages over insulin [14, 15]. Metformin, a bigua-
nide derivate used as first-line oral treatment for type 2 dia-
betes mellitus (T2DM) over 50 years, has been increasingly
used as an effective and safe agent in the treatment of GDM
[16–18]. Metformin treatment is superior to insulin in the
aspects of hepatic gluconeogenesis reduction, and peripheral
glucose uptake improvement without weight gain or hypo-
glycemia [19, 20]. Notably, insulin resistance, an important
feature of pregnancy and being more marked in gestational
diabetes, can be dealt more effectively with metformin [21,
22]. Indeed, a growing body of studies appeared to demon-
strate that metformin could be an effective and safe alter-
native or additional treatment to insulin for GDM women
[23–25]. Nevertheless, it has not yet been formally approved
by the US Food and Drug Administration for GDM treatment
[14], which may mainly attribute to the limited trial data,
especially in safety respect, and the discrepancies of clinical
outcomes across studies regarding metformin in GDM. To
provide evidence on the effectiveness and safety of metfor-
min in GDM, three meta-analyses were recently conducted
to compare the effects of oral hypoglycemic agents (OHAs)
(metformin or glyburide) with insulin in management of
GDM [26–28]. The study by Dhuliktia et al. failed to dem-
onstrate significant differences in glycemic control and
pregnancy outcomes between OHAs and insulin treatment.
Similarly, Su et al. also did not confirm significantly
increased adverse maternal and neonatal outcomes with the
use of metformin in GDM as compared with insulin treat-
ment. However, another recent meta-analysis including five
RCTs (1,270 patients) by Gui et al. showed that metformin
treatment had significant benefits for pregnant women with
GDM, in terms of average weight gains after enrollment,
incidence of pregnancy-induced hypertension (PIH), and
fasting blood glucose (FBG). Although these meta-analyses
appear to support the application of metformin treatment for
GDM, the included studies were limited. Besides, conflicting
results existed with some respects among these studies.
Furthermore, several RCTs comparing the effects of met-
formin with insulin for GDM have recently reported incon-
sistent results [25, 29, 30].
Therefore, we conducted this meta-analysis to compare
the effect of metformin with insulin on glycemic control,
maternal and neonatal outcomes in GDM, expecting to
provide sufficient and valid evidence for the application of
metformin in the management of GDM.
123
Arch Gynecol Obstet
Methods
Search strategy and selection criteria
Following the PRISMA statement guidelines for meta-
analysis of RCTs [31], we performed a literature search in
PUBMED, EMBASE, Science Direct, Springer link, and
Cochrane library databases from inception to May 31,
2014, without language restriction. The terms used to
search were ‘‘Gestational Diabetes’’ or ‘‘GDM’’ and
‘‘insulin’’ and ‘‘metformin’’. To identify potentially eligible
articles, the reference lists of published review articles and
included studies were manually searched.
Studies were included if they met the following crite-
ria: (1) Randomized control trials (RCTs) comparing the
effect of two treatments (metformin vs insulin) on
maternal and neonatal outcomes in patients with GDM.
(2) Studies offering related information on maternal and
neonatal outcomes. (3) Maternal outcomes were glyco-
hemoglobin A1c (HbA1c), FBG, PIH, weight gain,
cesarean delivery, preeclampsia, and gestational age;
neonatal outcomes were hypoglycemia, birth weight, large
for gestational age (LGA), small for gestational age
(SGA), premature birth, birth trauma, birth defect, neo-
natal intensive care unit (NICU), hyperbilirubinemia, and
shoulder dystocia.
The exclusion criteria were as follows: (1) Studies
published with insufficient information; (2) Reviews, let-
ters and comments were excluded; (3) Duplicate studies
were excluded. In the case that multiple publications were
from one study cohort, only the publication with the most
detailed information was included.
Data extraction and quality assessment
Two investigators independently extracted the following
data from the eligible studies using a standard protocol:
name of the first author, publication year, study region,
age and BMI of study subjects, sample size, medication
dosage, gestational age at randomization, criteria for
diagnosis and starting medical treatment of GDM, and
study outcomes. The discrepancies were resolved by
discussion. Quality assessment of studies was determined
with Quality Assessment of Diagnostic Accuracy Studies
(QUADAS) based on the following items: random
sequence generation, allocation concealment, blinding of
participants and personnel, blinding of outcome assess-
ment, incomplete outcome data, selective reporting, and
other bias [32]. Each item for corresponding study can be
defined as either low risk, high risk or unclear risk based
on the evaluation standard, and the overall bias can be
directly and objectively reflected by the combined results
of these items.
Arch Gynecol Obstet
Statistical analysis
Mean difference (MD) and 95 % confidence interval (CI)
were used for continuous outcomes; risk ratio (RR) and 95 %
CI were used for dichotomous outcomes. The heterogeneity
across studies was examined by the Cochran’s Q statistic and
the I2 statistic [33]. The fixed-effects model was selected for
homogeneous outcomes (P [ 0.05, I2 \ 50 %) and the
random-effects model was applied for heterogeneous out-
comes (P B 0.05, I2 C 50 %). The publication bias was
evaluated by funnel plot. A P value less than 0.05 was con-
sidered to be statistically significant. Statistical analysis was
performed with Review Manger version 5.2 software.
Results
Search results
The flow chart of study selection process is depicted in
Fig. 1. Under the search strategy, 2,108 articles were
retrieved, of which 580 were excluded because they were
duplicate publication. After screening the titles and
abstracts, 1,512 were excluded. With further examination,
another 5 of 16 articles were excluded (3 were non-RCTs, 1
was duplicate publication, 1 without available data for
analysis.) No additional eligible articles were obtained via
manual literature searches. Thus, 11 articles were finally
included in our meta-analysis [24, 25, 29, 30, 34–40].
Characteristics of included studies
The characteristics of included studies are shown in
Table 1 and the criteria used for diagnosis and criteria for
starting medical treatment are shown in Table 2. A total of
Fig. 1 Literature search and study selection
2,712 patients (1,354 in metformin group, and 1,358 in
insulin group) from 11 articles published from 2007 to
2014 were selected for present study. The result of quality
assessment is presented in Fig. 2. Since limited included
studies were designed with blind method, relatively high
risk was observed in performance bias and detection bias.
However, relatively low risk was found with respect to
overall bias, which suggested that the quality of the
included studies was moderate.
Primary maternal outcomes
Four primary maternal outcomes including HbA1c (at
gestational 36–37 weeks), FBG, PIH, and preeclampsia
were analyzed, and the fixed-effects model was used for all
these outcomes (P [ 0.05, I2 \ 50 %) besides HbA1c
(P \ 0.05, I2 [ 50 %). The pooled result revealed that
there was no significant difference between the two treat-
ments in HbA1c levels (Pheterogeneity = 0.0007, I2 = 82 %,
n = 959, MD = 0.06, 95 % CI -0.08 to 0.21, P = 0.37,
Fig. 3a), FBG (Pheterogeneity = 0.46, I2 = 0, n = 1,048,
MD = -0.32, 95 % CI -1.76 to 1.11, P = 0.66, Fig. 3b),
and the incidence of preeclampsia (Pheterogeneity = 0.69,
I2 = 0, n = 1,634, RR = 0.81, 95 % CI 0.55–1.17,
P = 0.26, Fig. 3d). On the other hand, significantly
reduced rate of PIH was found in the metformin group as
compared with the insulin group (Pheterogeneity = 0.71,
I2 = 0, n = 1,110, RR = 0.53, 95 % CI 0.31–0.90,
P = 0.02, Fig. 3c).
Secondary maternal outcomes
The fixed-effects model was used for all secondary
maternal outcomes, and the results showed that there was
no significant difference between the two treatments in
cesarean delivery rate (Pheterogeneity = 0.09, I2 = 44 %,
n = 1,594, RR = 0.94, 95 % CI 0.82–1.08, P = 0.36),
whereas significant decreased results were observed for
weight gain after enrollment (Pheterogeneity = 0.10,
I2 = 49 %, n = 1,530, MD = -1.28, 95 % CI -1.54 to
-1.01, P \ 0.0001) and gestational age at delivery
(Pheterogeneity = 0.26, I2 = 20 %, n = 2,072, MD = 0.94,
95 % CI -0.21 to -0.01, P = 0.03) in the metformin
treatment, compared with insulin treatment.
Primary neonatal outcomes
The fixed-effects model was used for all primary neonatal
outcomes, and the pooled analysis indicated a significantly
lower birth weight in the metformin group as compared
with the insulin group (Pheterogeneity = 0.30, I2 = 15 %,
n = 2,272, MD = -44.35, 95 % CI -85.79 to -2.90,
P = 0.04, Fig. 4a). There were seven studies which
123
123
Table 1 Characteristics of 11 included studies in this meta-analysis
References Area Age, metformin Age, insulin n1 n2 Dose (mg/ Dose (u) BMI, BMI, insulin Gestation weeks at Maternal Neonatal
d) metformin randomization outcomes outcomes

Moore et al. USA 27.1 (4.7) 27.7 (6.7) 32 31 1,000–2,000 NA NA NA 27.8 28.9 (5.0) d, e a, b, f, h, i
[40] (6.5)
Rowan et al. Multi- 33.5 (5.4) 33.0 (5.1) 363 370 1,750–2,500 50 35.1 (8.3) 34.6 (7.2) 30.2 30.1 (3.2) b, c, d, e, a, b, c, d,
[29] center (30–90) (3.3) f, g e, f, g,
i, j
Tertti et al. Finland 31.9 (5.0) 32.1 (5.4) 110 107 500–2,000 2–42 29.4 (5.9) 28.9 (4.7) 30.3 30.4 (1.8) a, c, d, e, a, b, c, e, f,
[39] (2.0) f, g g, i, j
Niromanesh Iran 30.7 (5.5) 31.8 (5.1) 80 80 1,000–2,500 NA 28.1 (4.0) 27.1 (2.1) 28.7 28.6 (3.6) a, b, c, d, a, b, c, d,
et al. [24] (3.7) e, f, g e, f, g, h,
i
Hasan et al. Pakistan 30.29 (3.06) 30.88 (3.6) 75 75 500–3,000 NA 29.17 (1.94) 28.74 (2.69) 29.53 29.2 (1.48) a, d, e a, b, d, h, i,
[30] (1.33) j
Ijäs et al. Finland 32.3 (5.6) 31.7 (6.1) 47 50 750–2,250 30 31.5 (6.5) 30.8 (5.4) 30.0 30.0 (4.0) d, e a, b, c, h, i,
[25] (4.9) j
Spaulonci Brazil 31.93 (6.02) 32.76 (4.66) 46 46 1,700–2,550 NA 31.96 (4.75) 31.39 (5.71) 32.18 32.05 (3.50) b, c, d, a, b, c, d,
et al. [38] (3.70) e, g e, h
Ibrahim et al. Egypt 29.84 (5.37) 29.84 (5.37) 46 44 1,500–2,000 1.14A 31.83 (3.23) 28.7 (3.71) d, e a, b, g, i
[36]
Barrett et al. Multi- 33.3 (32.6–34.0) 32.9 (32.2–33.5) 219 213 1,750–2,500 NA 34.5 33.75 30.1 30.3 a, c, g –
[34] center (33.5–35.5) (32.73–34.77) (29.7 (29.9–30.7)
–30.6)
Barrett et al. Multi- 33.1 (32.4–33.8) 32.5 (31.9–33.2) 236 242 1,750–2,500 NA 35.2 34.6 20–33 20–33 e, b, c, d
[35] center (34.2–36.2) (33.7–35.5)
Mesdaghinia Iran 29.6 (5.3) 30.2 (5.9) 100 100 500–2,500 0.5/kg/d 27.6 28.46 27.9 28.9 (3.8) – a, b, c, e,
et al. [37] (3.22) f, g
Multi-center: New Zealand and Australia; n1: case of Metformin group; n2: case of Insulin group
Maternal outcomes: a HbA1c at 36–37 weeks gestation, b fasting blood glucose, c weight gain after enrollment, d Cesarean delivery, e gestational age at delivery, f incidence of pregnancy-
induced hypertension, g incidence of preeclampsia
Neonatal outcomes: a incidence of hypoglycemia, b birth weight, c prevalence of large for gestational age infants, d prevalence of small for gestational age infants, e premature birth, f shoulder
dystocia, g birth defect, h incidence of hyperbilirubinemia, i admission to neonatal intensive care unit, j birth trauma
BMI body mass index
A
IU/kg
Arch Gynecol Obstet
Arch Gynecol Obstet

Table 2 Criteria for diagnosis Study Criteria for diagnosis of GDM Criteria for starting medical
and starting medical treatment treatment
of GDM
Loading Fasting 1h 2h 3h Fasting Postprandial
sugar (g) (mg/dl) (mg/dl) (mg/dl) (mg/dl) (mg/dl) (mg/dl)a

Moore et al. [40] 100 105 190 165 145 105 120
Rowan et al. [29] 75 99 126 97.2 120.6
Tertti et al. [39] 75 95.4 180 154.8 99 140.4 (1 h)
Niromanesh et al. [24] 50 95 130 100 95 120
Hasan et al. [30] 75 95 180 155 100 126
Ijäs et al. [25] 75 95.4 198 172.8 95.4 120.6
Spaulonci et al. [38] 75/100 95 180 155 140 95 120
GDM gestational diabetes
mellitus Ibrahim et al. [36] 50 95 130 100 95 120
a
The postprandial levels (2 h) Barrett et al. [34] 50 95 130 100 95 120
offered in most included studies, Barrett et al. [35] 50 95 130 100 95 120
and postprandial levels (1 h) Mesdaghinia et al. [37] 100 95 180 155 140 95 120
offered in two included studies

provided the data on LGA infants; meta-analysis of these
data failed to demonstrate a significant difference between
the two treatments (Pheterogeneity = 0.23, I2 = 26 %,
n = 1,977, RR = 0.91, 0.75–1.10, P = 0.33, Fig. 4b).
Similarly, no significant difference was indicated according
to the pooled analysis on the rate of SGA infants between
the two treatments (Pheterogeneity = 0.49, I2 = 0,
n = 1,710, RR = 0.88, 95 % CI 0.66–1.20, P = 0.43,
Fig. 4c). Hypoglycemia was reported in nine studies, and
the combined result showed a significantly lower rate of
hypoglycemia in the metformin group as compared with
the insulin group (Pheterogeneity = 0.28, I2 = 18 %,
n = 1,793, RR = 0.69, 95 % CI 0.55–0.87, P = 0.001,
Fig. 4d). As shown in the pooled result of six studies, there
was no significant difference between the two treatments
with respect to the incidence of hyperbilirubinemia
(Pheterogeneity = 0.36, I2 = 8 %, n = 778, RR = 0.88,
95 % CI 0.63–1.22, P = 0.43, Fig. 4e).
Secondary neonatal outcomes
The fixed-effects model was used for all secondary
neonatal outcomes, and meta-analysis revealed that there
was no significant difference between the two treatments
in premature birth (Pheterogeneity = 0.17, I2 = 38 %,
n = 1,402, RR = 1.32, 95 % CI 0.92–1.88, P = 0.13),
shoulder dystocia (Pheterogeneity = 0.58, I2 = 0,
n = 1,373, RR = 0.73, 95 % CI 0.36–1.47, P = 0.37),
birth defect (Pheterogeneity = 0.52, I2 = 0, n = 1,382,
RR = 0.76, 95 % CI 0.44–1.31, P = 0.32), and birth
trauma (Pheterogeneity = 0.39, I2 = 0, n = 1,197,
RR = 0.79, 95 % CI 0.42–1.46, P = 0.39). However,
compared with the insulin group, significantly decreased
rate of NICU was observed in the metformin group
(Pheterogeneity = 0.49, I2 = 0, n = 1,501, RR = 0.82,
95 % CI 0.67–0.99, P = 0.04).
Publication bias
Publication bias was assessed for the outcome of birth
weight which was reported in most included studies. As
shown in Fig. 5, no obvious publication bias was observed
with the relatively symmetrical funnel plots.
Discussion
A total of 11 studies comprising 2,712 patients were
included for present meta-analysis. With respect to the
primary maternal outcomes, there was no significant dif-
ference between the two groups except that PIH rate was
significantly lower in the metformin group. The pooled
analysis regarding secondary outcomes of pregnant women
revealed that when compared with the insulin group, sig-
nificantly lower average weight gains after enrollment and
average gestational ages at delivery existed in the metfor-
min group. This finding is consistent with the previous
meta-analysis [26], and suggests a few benefits and risks of
metformin treatment for GDM.
Although the cellular and molecular mechanisms of
metformin remain unclear, Viollet et al. [41] considered
that the benefit from metformin in PIH rate may be
attributed to its effect on endothelial functions and reactive
oxygen species production. In fact, meta-analyses have
confirmed a reduced effect on cardiovascular mortality
with metformin, as compared with any other oral diabetes
agents [42, 43]. In the current study, similar HbA1c and

123

Fig. 2 Assessment of the risk of bias in the included studies.
a Methodological quality graph: authors’ judgment about each
methodological quality item presented as percentages across all
included studies; b methodological quality summary: authors’
FBG levels were observed between the two treatments,
which suggested that metformin was comparable with
insulin in glycemic control. Though the anti-hyperglycemic
mechanism of metformin required to be confirmed, it was
found that the liver was the main site of action for met-
formin, and suppression of hepatic gluconeogenesis was
regarded as the most widely accepted model of the anti-
hyperglycemic action of metformin [44].
On the other hand, our results indicated that the average
birth weights, as well as the incidence of hypoglycemia and
NICU were significantly lower in the metformin group as
compared with the insulin group. This result, however,
differs from the meta-analysis by Gui et al. [26] which
showed there was no significant difference between the two
groups in the rate of hypoglycemia or NICU. In addition,
they failed to demonstrate significant lower average birth
weights in the metformin group, although slightly lower
tendency was found. Similarly, the other meta-analysis also
123
Arch Gynecol Obstet
judgment about each methodological quality item for each included
study. ?: low risk of bias; ‘‘?’’: unclear risk of bias; ‘‘-’’: high risk of
bias
could not confirm significant difference between the two
groups in NICU or birth weights; whereas significant lower
incidence of hypoglycemia was presented for the metfor-
min group [27]. It is noteworthy that all 10 included studies
regarding birth weights did not find a significant difference
between the two groups. We considered that the pooled
result of our study regarding birth weights may be influ-
enced by some potential bias. The more recent RCT of
Ibrahim et al. [36] demonstrated a significantly reduced
rate of NICU and hypoglycemia in the metformin group.
Similar results were obtained in the study of Hasan et al.
[30], and a rational interpretation suggested that signifi-
cantly lower neonatal NICU admissions were associated
with lesser hypoglycemia and hyperbilirubinemia at birth
in metformin group. However, in the study of Rowan et al.
[29] similar rates of neonatal hypoglycemia were found
between metformin and insulin groups, although the result
observed a higher proportion of severe neonatal
Arch Gynecol Obstet

Fig. 3 Forest plots of primary maternal outcomes comparing metformin with insulin. a HbA1c, b FBG, c incidence of PIH, d incidence of
preeclampsia. MD mean differences, CI confidence intervals, RR risk ratio, FBG fasting blood glucose, PIH pregnancy-induced hypertension

hypoglycemia (glucose less than 28.8 mg/dL) in the insulin
group (8.1 %) compared with the metformin group (3.3 %,
P = 0.008). The authors considered that metformin could
cross the placenta to cause neonatal hypoglycemia, but was
less likely to cause severe neonatal hypoglycemia [29].
Actually, metformin crosses the placenta with a maternal-
to-fetal transfer rate estimated of 10–16 %, which may
naturally affect fetal physiology directly [45]. In the cur-
rent meta-analysis, the finding that a significantly reduced
rate of hypoglycemia in the metformin group requires to be
validated because of limited available information in this
regard to date, meanwhile, the exact mechanism by which
metformin affect the fetal physiology remains to be
elucidated.
All 11 included studies for present meta-analysis are
RCTs with relatively high quality. In addition, there was
no significant heterogeneity across studies for all study
outcomes except for HbA1c, and no obvious publication
bias was observed. On the other hand, potential limita-
tions should be considered for our result. First, limited

123
 Arch Gynecol Obstet

Fig. 4 Forest plots of primary

neonatal outcomes comparing
metformin with insulin. a Birth
weight, b incidence of SGA
infants, c incidence of LGA
infants, d incidence of
hypoglycemia, e incidence of
hyperbilirubinemia. MD mean
differences, CI confidence
intervals, RR risk ratio, LGA
large for gestational age, SGA
small for gestational age

123
Arch Gynecol Obstet
Fig. 5 Funnel plots for publication bias analysis
data were available with respect to some outcomes
including PIH, HbA1c, and FBG, which may mask the
facticity of effect value. Second, most of the included
studies were not designed with blind method, which
caused relatively high risk of performance bias and
detection bias, and this may influence the overall result.
Therefore, further studies with sufficient power to detect
meaningful differences in maternal and neonatal out-
comes are warranted.
Conclusion
Metformin can significantly reduce several adverse
maternal and neonatal outcomes including PIH rate, inci-
dence of hypoglycemia and NICU, and it may be an
effective and safe treatment for the women with GDM.
However, potential risk of metformin in the management of
GDM is worthy further investigation.
Conflict of interest The authors declare that they have no conflicts
of interest to disclose.
References
1. Association AD (2003) Gestational diabetes mellitus. Diabetes
care 26:S103
2. Ferrara A (2007) Increasing prevalence of gestational diabetes
mellitus a public health perspective. Diabetes Care 30(Supple-
ment 2):S141–S146
3. Coustan DR (2013) Gestational diabetes mellitus. Clin Chem
59(9):1310–1321
4. Jenum A K, Mørkrid K, Sletner L, Vange S, Torper J L, Nakstad B,
Voldner N et al (2012) Impact of ethnicity on gestational diabetes
identified with the WHO and the modified international association
of diabetes and pregnancy study groups criteria: a population-based
cohort study[J]. Eur J Endocrinol 166(2):317–324
5. Catalano PM, McIntyre HD, Cruickshank JK, McCance DR,
Dyer AR, Metzger BE, Lowe LP, Trimble ER, Coustan DR,
Hadden DR (2012) The hyperglycemia and adverse pregnancy
outcome study associations of GDM and obesity with pregnancy
outcomes. Diabetes Care 35(4):780–786
6. Deveer R, Deveer M, Akbaba E, Engin-Ustun Y, Aydogan P,
Celikkaya H, Danisman N, Mollamahmutoglu L (2013) The
effect of diet on pregnancy outcomes among pregnant with
abnormal glucose challenge test. Eur Rev Med Pharmacol Sci
17(9):1258–1261
7. Barakat R, Pelaez M, Lopez C, Lucia A, Ruiz JR (2013) Exercise
during pregnancy and gestational diabetes-related adverse effects:
a randomised controlled trial. Br J Sports Med 47(10):630–6
8. Evensen AE (2012) Update on gestational diabetes mellitus. Prim
Care 39(1):83–94
9. Nicholson W, Baptiste-Roberts K (2011) Oral hypoglycaemic
agents during pregnancy: the evidence for effectiveness and
safety. Best Practi Res Clin Obstet Gynaecol 25(1):51–63
10. Glueck CJ, Goldenberg N, Streicher P, Wang P (2002) The
contentious nature of gestational diabetes: diet, insulin, glyburide
and metformin. Expert Opin Pharmacother 3(11):1557–1568
11. Association AD (2012) Standards of medical care in diabetes—
2012. Diabetes Care 35:S11
12. Norman RJ, Wang JX, Hague W (2004) Should we continue or
stop insulin sensitizing drugs during pregnancy? Curr Opin
Obstet Gynecol 16(3):245–250
13. Petry CJ (2010) Gestational diabetes: risk factors and recent
advances in its genetics and treatment. Br J Nutr 104(06):775–787
14. Homko CJ, Reece EA (2006) Insulins and oral hypoglycemic agents
in pregnancy. J Matern Fetal Neonatal Med 19(11):679–686
15. Giugliano E, Cagnazzo E, Giugliano B, Caserta D, Moscarini M,
Marci R (2013) The prevention of gestational diabetes. J Diabetes
Metab 4(7):286
16. Heilmaier C, Thielscher C, Ziller M, Altmann V, Kostev K
(2014) Use of antidiabetic agents in the treatment of gestational
diabetes mellitus in Germany, 2008–2012. J Obstet Gynaecol Res
40(6):1592–1597
17. Inzucchi S, Bergenstal R, Buse J, Diamant M, Ferrannini E,
Nauck M, Peters A, Tsapas A, Wender R, Matthews D (2012)
Management of hyperglycaemia in type 2 diabetes: a patient-
centered approach. Position statement of the American Diabetes
Association (ADA) and the European Association for the Study
of Diabetes (EASD). Diabetologia 55(6):1577–1596
18. Nathan DM, Buse JB, Davidson MB, Ferrannini E, Holman RR,
Sherwin R, Zinman B (2009) Medical management of hyper-
glycemia in type 2 diabetes: a consensus algorithm for the initi-
ation and adjustment of therapy a consensus statement of the
American Diabetes Association and the European Association for
the Study of Diabetes. Diabetes Care 32(1):193–203
19. Wensel TM (2009) Role of metformin in the treatment of ges-
tational diabetes. Ann Pharmacother 43(5):939–943
20. Dasari P, Habeebullah S Maternal and fetal outcome in gesta-
tional diabetes mellitus (GDM) treated with diet and metformin-a
preliminary retrospective analysis. Open Conf Proc J 2:59–63
21. Evans JL, Youngren JF, Goldfine ID (2004) Effective treatments
for insulin resistance: trim the fat and douse the fire. Trends
Endocrinol Metab 15(9):425–431
22. Giannarelli R, Aragona M, Coppelli A, Del Prato S (2003)
Reducing insulin resistance with metformin: the evidence today.
Diabetes Metab 29(4):6S28–26S35
23. Marques P, Carvalho MR, Pinto L, Guerra S (2014) Metformin
safety in the management of gestational diabetes. Endocr Pract
1(1):1–21
24. Niromanesh S, Alavi A, Sharbaf FR, Amjadi N, Moosavi S,
Akbari S (2012) Metformin compared with insulin in the man-
agement of gestational diabetes mellitus: a randomized clinical
trial. Diabetes Res Clin Pract 98(3):422–429
25. Ijäs H, Vääräsmäki M, Morin-Papunen L, Keravuo R, Ebeling T,
Saarela T, Raudaskoski T (2011) Metformin should be considered
123

in the treatment of gestational diabetes: a prospective randomised
study. BJOG 118(7):880–885
26. Gui J, Liu Q, Feng L (2013) Metformin vs insulin in the man-
agement of gestational diabetes: a meta-analysis. PLOS One
8(5):e64585
27. Su D, Wang X (2014) Metformin vs insulin in the management of
gestational diabetes: a systematic review and meta-analysis.
Diabetes Res Clin Pract 104(3):353-357
28. Dhulkotia JS, Ola B, Fraser R, Farrell T (2010) Oral hypogly-
cemic agents vs insulin in management of gestational diabetes: a
systematic review and metaanalysis. Am J Obstet Gynecol
203(5):457–459
29. Rowan JA, Hague WM, Gao W, Battin MR, Moore MP (2008)
Metformin versus insulin for the treatment of gestational diabe-
tes. N Engl J Med 358(19):2003–2015
30. Hasan JA, Karim N, Sheikh Z (2012) Metformin prevents mac-
rosomia and neonatal morbidity in gestational diabetes. Pak J
Med Sci 28(3):384–389
31. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC,
Ioannidis JP, Clarke M, Devereaux P, Kleijnen J, Moher D (2009)
The PRISMA statement for reporting systematic reviews and
meta-analyses of studies that evaluate health care interventions:
explanation and elaboration. Ann Intern Med 151(4):W-65–W-94
32. Higgins JP, Green S, Collaboration C (2008) Cochrane handbook
for systematic reviews of interventions, vol 5. Wiley Online
Library, Europe
33. Higgins JP, Thompson SG, Deeks JJ, Altman DG (2003) Mea-
suring inconsistency in meta-analyses. BMJ 327(7414):557
34. Barrett HL, Gatford KL, Houda CM, De Blasio MJ, McIntyre
HD, Callaway LK, Nitert MD, Coat S, Owens JA, Hague WM
(2013) Maternal and neonatal circulating markers of metabolic
and cardiovascular risk in the metformin in gestational diabetes
(MiG) trial responses to maternal metformin versus insulin
treatment. Diabetes Care 36(3):529–536
35. Barrett HL, Nitert MD, Jones L, O’Rourke P, Lust K, Gatford
KL, De Blasio MJ, Coat S, Owens JA, Hague WM (2013)
Determinants of maternal triglycerides in women with gestational
diabetes mellitus in the metformin in gestational diabetes (MiG)
study. Diabetes Care 36(7):1941–1946
123
Arch Gynecol Obstet
36. Ibrahim MI, Hamdy A, Shafik A, Taha S, Anwar M, Faris M
(2014) The role of adding metformin in insulin-resistant diabetic
pregnant women: a randomized controlled trial. Arch Gynecol
Obstet 289(5):959–965
37. Mesdaghinia E, Samimi M, Homaei Z, Saberi F, Moosavi SGA,
Yaribakht M (2013) Comparison of newborn outcomes in women
with gestational diabetes mellitus treated with metformin or
insulin: A randomised blinded trial. Int J Prev Med 4(3):327
38. Spaulonci CP, Bernardes LS, Trindade TC, Zugaib M, Francisco
RPV (2013) Randomized trial of metformin vs insulin in the
management of gestational diabetes. Am J Obstet Gynecol
209(1):34. e31–34. e37
39. Tertti K, Ekblad U, Koskinen P, Vahlberg T, Rönnemaa T (2013)
Metformin vs. insulin in gestational diabetes. A randomized study
characterizing metformin patients needing additional insulin.
Diabetes Obes Metab 15(3):246–251
40. Moore LE, Briery CM, Clokey D, Martin RW, Williford NJ,
Bofill JA, Morrison JC (2007) Metformin and insulin in the
management of gestational diabetes mellitus: preliminary results
of a comparison. J Reprod Med 52(11):1011–1015
41. Viollet B, Guigas B, Garcia NS, Leclerc J, Foretz M, Andreelli F
(2012) Cellular and molecular mechanisms of metformin: an
overview. Clin Sci 122(6):253–270
42. Lamanna C, Monami M, Marchionni N, Mannucci E (2011)
Effect of metformin on cardiovascular events and mortality: a
meta-analysis of randomized clinical trials. Diabetes Obes Metab
13(3):221–228
43. Selvin E, Bolen S, Yeh H-C, Wiley C, Wilson LM, Marinopoulos
SS, Feldman L, Vassy J, Wilson R, Bass EB (2008) Cardiovas-
cular outcomes in trials of oral diabetes medications: a systematic
review. Arch Inter Med 168(19):2070–2080
44. Rena G, Pearson ER, Sakamoto K (2013) Molecular mechanism
of action of metformin: old or new insights? Diabetologia
56(9):1898–1906
45. Nanovskaya TN, Nekhayeva IA, Patrikeeva SL, Hankins GD,
Ahmed MS (2006) Transfer of metformin across the dually per-
fused human placental lobule. Am J Obstet Gynecol 195(4):
1081–1085