DIABET-702;

+Model No. of Pages 9

ARTICLE IN PRESS
Available online at

ScienceDirect
www.sciencedirect.com
Diabetes & Metabolism xxx (2015) xxxxxx

Original article

Pregnancy adverse outcomes related to pregravid body mass index and

gestational weight gain, according to the presence or not of gestational
diabetes mellitus: A retrospective observational study
E. Cosson a,b, , C. Cussac-Pillegand a , A. Benbara c , I. Pharisien c , M.T. Nguyen a , S. Chiheb
a
, P. Valensi a , L. Carbillon c
a

Department of Endocrinology-Diabetology-Nutrition, CRNH-IdF, CINFO, Paris 13 University, Sorbonne Paris Cit, Jean-Verdier Hospital, APHP,
Bondy, France
b
Sorbonne Paris Cit, UMR U1153 Inserm, U1125 Inra, Cnam, Universit Paris 13, Bobigny, France
c
Department of Obstetrics and Gynecology, Paris 13 University, Sorbonne Paris Cit, Jean-Verdier Hospital, APHP, Bondy,
France
Received 6 February 2015; received in revised form 1st June 2015; accepted 2 June 2015

Abstract
Aim. This study retrospectively evaluated the complications associated with prepregnancy overweight (OW) or obesity (OB) and gestational
weight gain (GWG) in women with or without universally screened and treated gestational diabetes mellitus (GDM).
Methods. A total of 15,551 non-Asian women without pregravid diabetes or hypertension who delivered singleton babies (20022010)
were classified according to GDM (13.5%), pregestational body mass index (BMI; normal range: 18.524.9 kg/m2 ), OW (26.2%), OB (13.9%;
BMI 30 kg/m2 ) and GWG (< 7 kg: 32%; 711.5 kg: 37%; 11.616 kg: 23%; > 16 kg: 8%). Main outcome measures were large/small for
gestational age (LGA/SGA), caesarean section, preeclampsia, preterm delivery and shoulder dystocia.
Results. GDM was associated with more LGA babies [Odds Ratio (OR): 2.12, 95% confidence interval (CI): 1.852.43], caesarean section
(OR: 1.49, 95% CI: 1.341.65) and preeclampsia (OR: 1.59, 95% CI: 1.212.09). OW/OB and GWG were associated with LGA infants
whatever the GDM status, and with SGA babies only in women without GDM. LGA status was independently associated with GWG in women
with GDM (11.616 kg: OR: 1.74, 95% CI: 1.492.03 and > 16 kg OR: 3.42, 95% CI: 2.834.13 vs 711.5 kg) and in women without GDM
(OR: 2.14, 95% CI: 1.542.97 or OR: 2.65, 95% CI: 1.684.17, respectively), and with BMI only in women without GDM (OR: 1.12, 95%
CI: 1.001.24, per
10 kg/m2 ). SGA status was independently associated with OW (OR: 0.86, 95% CI: 0.770.98), OB (OR: 0.84, 95% CI: 0.720.98) and GWG < 7
kg
(1.14, 95% CI: 1.011.29) only in women without GDM.
Conclusion. In our European cohort and considering the triumvirate of GDM, BMI and GWG, GDM was the main contributor to caesarean
section and preeclampsia. OW/OB and GWG contributed to LGA and SGA infants mainly in women without GDM.
2015 Elsevier Masson SAS. All rights reserved.
Keywords: Gestational diabetes mellitus; Gestational weight gain; Obesity; Pregnancy; Prognosis

1. Introduction
Abbreviations: GWG, Gestational weight gain; IOM, Institute of
Medicine; HAPO, Hyperglycaemia and Adverse Pregnancy Outcomes;
IADPSG, Interna- tional Association of Diabetes and Pregnancy Study
Groups; LGA, large for gestational age; SGA, small for gestational age.

Corresponding author at: Department of Endocrinology-DiabetologyNutrition, hpital Jean-Verdier, avenue du 14-juillet, 93143 Bondy cedex,
France. Tel.: +33 148 02 65 80; fax: +33 148 02 65 79.
E-mail address: emmanuel.cosson@jvr.aphp.fr (E. Cosson).

Gestational diabetes mellitus (GDM) is defined as any

degree of glucose intolerance with onset or first recognition
during pregnancy, and is associated with adverse outcomes
during pregnancy [1]. Obesity has a growing prevalence in
women of childbearing age [2] and is a confounding factor.
First, it is a risk factor for GDM [2,3]. Second, it shares
complications with GDM, such as large-for-gestational-age
(LGA) infants

Please cite this article in press as: Cosson E, et al. Pregnancy adverse outcomes related to pregravid body mass index and gestational
weight gain, according to the presence or not of gestational diabetes mellitus: A retrospective observational study. Diabetes Metab (2015),
http://dx.doi.org/10.1016/j.diabet.2015.06.001

DIABET-702;
+Model No. of Pages 9

ARTICLE IN PRESS

http://dx.doi.org/10.1016/j.diabet.2015.06.001
1262-3636/ 2015 Elsevier Masson SAS. All rights reserved.

Please cite this article in press as: Cosson E, et al. Pregnancy adverse outcomes related to pregravid body mass index and gestational
weight gain, according to the presence or not of gestational diabetes mellitus: A retrospective observational study. Diabetes Metab (2015),
http://dx.doi.org/10.1016/j.diabet.2015.06.001

E. Cosson et al. / Diabetes & Metabolism xxx (2015) xxxxxx

[411], caesarean section [4,5,7,8,11,12], hypertensive disorders [4,5,7,8] and, in certain studies, shoulder dystocia [5].
Also, gestational weight gain (GWG) appears to be crucial
[5,810,13,14].
To date, only five recent studies, four from the United
States [5,10,15,16] and only one from Europe [9], have
explored the impact of GDM, obesity and GWG together.
Some limitations may affect these observational studies.
First, the prevalence of GDM is sometimes very low [9,16]
with screening which might not have been universal
[5,10,15,16]. Second, women with pregravid diabetes and
hypertension were not excluded [5,9,10,15,16], whereas these
conditions are often associated with overweight and obesity.
Therefore, considering women with isolated obesity might
better evaluate the role of obesity per se [12]. Regarding body
mass index (BMI), underweight women are not always
considered separately [5,16] nor is the lower BMI cutoff point
in Asian women [17] taken into account to define overweight
and obesity [5,9]. Finally, excessive GWG [9,10], determined
according to pregravid BMI status as pro- posed by the
Institute of Medicine (IOM) [18] rather than GWG per se, has
often been considered and is an additional confound- ing
factor.
Dietary advice and drugs are generally provided only to
women with GDM, as GWG [5,810,13,14], treatment modalities and glycaemic levels achieved can modify the outcomes
[19]. Only the Hyperglycaemia and Adverse Pregnancy Outcomes (HAPO) study reported obesity-related adverse events
independently of glycaemic status and its treatment [4,7].
How- ever, in that study, BMI was measured at the time of
oral glucose tolerance tests at between 24 and 32 weeks of
gestation, and not before pregnancy. Therefore, GWG could
not be assessed.
Given this context, a large multiethnic European cohort of
non-Asian women who delivered singleton babies and were
without pregravid diabetes or hypertension was selected for
our present retrospective observational study. In this cohort,
the adverse outcomes related to isolated overweight, obesity
and GWG were investigated in women with and without
universally screened and treated GDM.
2. Methods
2.1. Participants, GDM screening and
care
A total of 20,653 women delivered at our hospital between
January 2002 and December 2010. Data are routinely entered
at birth for all women (no exceptions) giving birth at our
university hospital by the midwife assisting at the delivery,
then checked and collected during the maternity stay by a
midwife quali- fied in data management and storage (I.P.),
with no interactions with the women themselves. The authors
did not have access to identification of patients information
prior to anonymization. The purposes of the database are to
assess the overall quality of obstetric care and to regularly
update medical management protocols. The data are
retrospective and observational, with no need for either
approval by an ethics committee/institutional review board or
patients written informed consent. The patients

records/information are anonymous, and the database is

declared

to the French data protection authority (Commission nationale

de linformatique et des liberts [CNIL]).
In the present study, women with known diabetes (n = 204),
previous hypertensive disorders (n = 448) and multiple
pregnan- cies (n = 378) were not included. Furthermore,
women whose prepregnancy BMI (n = 1669) and GWG (n =
2) were unknown were also not included. Finally, those also
excluded were women from Asia (n = 628) or
India/Pakistan/Sri Lanka (n = 1076), and those with a BMI <
18.5 kg/m2 (n = 687).
Thus, 15,551 pregnancies were analyzed. Definitions of our
parameters did not change over the 9 years of the study. BMI
was calculated from self-reported pregravid weight and
measured height during pregnancy, using the following
formula: weight (in kg) divided by the height (in m) squared.
Women were clas- sified as normal weight, overweight and
obese when their BMI
was 18.524.9 kg/m2 , 25.029.9 kg/m2 and 30 kg/m2 ,
respectively [17]. GWG categories (< 7 kg, 7.011.5 kg, 11.616 kg
and > 16 kg) were defined according to the usual thresholds
proposed by IOM guidelines for overweight women (optimal
GWG: 711.5 kg) and normal-weight women (optimal GWG:
11.616 kg) [18].
GDM was assessed using a one-step screening and diagnostic test, which always comprised a 75-g oral glucose tolerance
test [2,20,21]. GDM was defined as a fasting plasma glucose
value 5.3 mmol/L (the same fasting plasma glucose target
as in previous French recommendations) and/or a 2-h blood
glu- cose value 7.8 mmol/L (World Health Organization
criteria) [2,20,21]. One-step screening was chosen to limit the

number of participants lost to follow-up, as our study

population was cha- racterized by widespread geographical
origins [21]. Screening was specifically prescribed during the
hospital routine follow-up visit and then performed out of
hospital. As is usual for epidemi- ological studies, the women
without screening were considered to be without GDM
[2,9,10,15].
Women who were overweight or obese had no specific
follow-up unless they were diagnosed with GDM. All women
with such a diagnosis were referred to a multidisciplinary
team, which included a diabetologist, obstetrician, midwife,
dietitian and nurse educator. These women received individualized dietary advice, were instructed on how to perform
self-monitoring of blood glucose levels six times a day, and
visited the diabetologist every two to four weeks. Insulin therapy was started if fasting and 2-h postprandial glucose levels
were > 5.3 mmol/L and > 6.8 mmol/L, respectively.
Antenatal visits were scheduled for every two to four weeks up
to 34 weeks, and weekly thereafter, with cardiotocography and
assessment of amniotic fluid volume [2,21].
2.2. Prognosis
The following outcomes were considered: LGA or SGA
(birth weight > 90th percentile or < 10th percentile,
respectively, of the general French population) [22]; caesarean
section; preeclampsia (blood pressure 140/90 mmHg on two
measure- ments taken 4 h apart and proteinuria 300 mg/24
h or 3+ or more on dipstick testing of a random urine
sample); preterm delivery (before 37 full weeks); and shoulder
dystocia, defined

E. Cosson et al. / Diabetes & Metabolism xxx (2015) xxxxxx

Table 1
Maternal characteristics and complications by gestational diabetes mellitus (GDM) status and pregravid body mass index (BMI).

Characteristics
Pregravid BMI (kg/m2 )
BMI classification
Normal weight (%)
Overweight (%)
Obesity (%)
Age (years)
Parity (n)
Multiparity (%)
Maternal smoking
Before pregnancy (%)
During pregnancy (%)
Ethnicity
Caucasian (%)
Sub-Saharan African (%)
Caribbean (%)
Other (%)
Family history of diabetes (%)
Previous pregnancy with macrosomia (%)
Previous pregnancy with GDM (%)
Events
GDM (%)
Gestational weight gain (kg)
Gestational weight gain classification
< 7 kg (%)
711.5 kg (%)
11.616 kg (%)
> 16 kg (%)
Large-for-gestational-age babies (%)
Small-for-gestational-age babies (%)
Caesarean section (%)
Preeclampsia (%)
Preterm delivery (%)
Shoulder dystocia (%)

Total cohort No GDM

GDM
(n = 15,551) (n = 13,436) (n = 2097)

ANOVA, Normal
weight
P
(n = 9317)

Overweight
(n = 4075)

Obesity
(n = 2159)

ANOVA,
P

24.6 4.7

24.6 4.7

24.9 4.8

< 0.001
< 0.01

21.6 1.6

26.6 1.4a

33.6 4.0a , b

< 0.001

9317 (59.9)
4075 (26.2)
2159 (13.9)
29.7 5.8
2.1 1.3
9045 (58.2)

8127 (60.4)
3489 (25.9)
1838 (13.7)
29.6 5.8
2.0 1.3
7728 (57.4)

1190 (56.7)
586 (27.9)
321 (15.3)
30.6 5.8
2.2 1.4
1317 (62.8)

29.7 5.9
2.0 1.2
5315 (57.0)

29.9 5.8
2.1 1.3a
2426 (59.5)a

29.7 5.9
2.2 1.4a
1304 (60.4)a

2243 (14.4)
1503 (9.7)

1994 (14.8)
1352 (10.0)

249 (11.9)
151 (7.2)

1421 (15.3)
946 (10.2)

526 (12.9)a
344 (8.4)a

296 (13.7)
213 (9.9)

9881 (63.5)
3566 (22.9)
1365 (8.8)
739 (4.8)
3227 (20.8)
457 (2.9)
498 (3.2)

8434 (62.7)
3181 (23.6)
1196 (8.9)
643 (4.8)
2718 (20.2)
366 (2.7)
268 (2.0)

1447 (69.0)
385 (18.4)
169 (8.1)
96 (4.6)
509 (24.3)
91 (4.3)
230 (11.0)

5998 (64.4)
2058 (22.1)
817 (8.8)
444 (4.8)
1950 (20.9)
266 (2.9)
276 (3.0)

2530 (62.1)
1020 (25.0)
324 (8.0)
201 (4.9)
833 (20.4)
124 (3.0)
136 (3.3)

1353 (62.7)
488 (22.6)
224 (10.4)
94 (4.4)
444 (20.6)
27 (3.1)
86 (4.0)a

2097 (13.5)
8.9 5.7

9.0 5.7

8.5 5.5

1190 (12.8)
9.1 5.6

586 (14.4)a
8.8 5.6

321 (14.9)a
8.5 5.8a

5041 (32.4)
5695 (36.6)
3586 (23.1)
1229 (7.9)
1341 (8.6)
2114 (13.6)
3265 (21.0)
335 (2.2)
1207 (7.8)
231 (1.5)

4272 (31.8)
4945 (36.8)
3149 (23.4)
1088 (8.1)
1028 (7.6)
1837 (13.7)
2696 (20.0)
269 (2.0)
1032 (7.7)
193 (1.4)

769 (36.7)
750 (35.8)
437 (20.8)
141 (6.7)
313 (14.9)
277 (13.1)
569 (27.1)
66 (3.1)
175 (8.3)
38 (1.8)

2902 (31.1)
3481 (37.4)
2180 (23.4)
754 (8.1)
778 (8.4)
1332 (14.3)
1944 (20.9)
185 (2.0)
733 (7.9)
137 (1.5)

1362 (33.4)
1471 (36.1)
929 (22.8)
313 (7.7)
361 (8.9)
517 (12.7)a
863 (21.2)
88 (2.2)
296 (7.3)
59 (1.4)

777 (32.4)
743 (29.6)
477 (22.1)
162 (7.5)
202 (9.4)
123 (13.9)a
458 (21.2)
62 (2.9)a
178 (8.2)
35 (1.6)

< 000.1
< 000.1
< 000.1
< 000.1
< 000.1
< 000.1

< 000.1
< 000.1
< 000.1

< 000.1
< 0.01

< 000.1
NS
< 000.1
< 000.1
NS
NS

NS
< 0.001
< 0.01
< 0.001
< 0.01
< 0.01

NS
NS
< 0.05
< 0.01
< 0.001
< 0.001

NS
< 0.01
NS
< 0.05
NS
NS

Data are presented as means SD or as n (%); NS: not significant.

a P < 0.05 vs women with BMI 18.524.9 kg/m2 .
b
P < 0.05 vs women with BMI 2529.9 kg/m2 .

as the use of obstetric manoeuvres (such as a McRoberts episiotomy after delivery of the fetal head, suprapubic pressure,
posterior arm rotation to an oblique angle, rotation of the
infant by 180 degrees and delivery of the posterior arm) [23].

SPSS software (SPSS, Chicago, IL, USA). The 0.05

probability level was considered statistically significant.
3. Results

2.3. Statistical analyses

3.1. Characteristics of the study population

Continuous variables were expressed as means SD,

and compared by one-way analysis of variance (ANOVA) or
the MannWhitney U test as adequate. The significance of
differ- ences in proportions was tested with the Chi2 test.
Logistic regression was used for analyses of BMI and GWG
effects on LGA and SGA infants, caesarean section and
preeclampsia in women with and without GDM. Logistic
regression was also used for multivariate analyses based on a
model including factors associated with LGA and then SGA,
with a P value < 0.10 on uni- variate analyses. All statistical
analyses were carried out using

Maternal characteristics are shown in Table 1. In summary,

the women were 29.7 5.8 years old, and their mean
parity was 2.1 1.3. GDM was diagnosed in 13.5%. The
mean pre- gravid BMI was 24.6 4.7 kg/m2 , with
overweight and obesity observed in 26.2% and 13.9%,
respectively. Mean GWG was
8.9 5.7 kg. Of note, the cohort was multiethnic, with most
of the subjects being Caucasian (from Europe or North
Africa;
63.5%) or from sub-Saharan Africa (22.9%).
Pregravid parameters associated with GDM were BMI, age,
parity, maternal smoking, ethnicity, family history of diabetes

E. Cosson et al. / Diabetes & Metabolism xxx (2015) xxxxxx

and previous pregnancy with macrosomia or GDM (Table 1).

Classes of increasing BMI (normal weight, overweight and
obe- sity) were associated with higher parity, less smoking
before and during pregnancy, ethnicity and a more frequent
personal history of GDM. Mean GWG was lower in obese
women (Table 1).
3.2. Pregnancy-related events associated with GDM
and overweight/obesity
Table 1 also shows that GDM was associated with less
GWG, and more LGA infants (OR: 2.12, 95% CI: 1.85
2.43), cae- sarean section (OR: 1.49, 95% CI: 1.341.65) and
preeclampsia (OR: 1.59, 95% CI: 1.212.09). An increased
BMI classification was associated with lower GWG, more
GDM, preeclampsia and fewer SGA infants (Table 1).
3.3. Complications associated with BMI and GWG
according
to
GDM
status
On analyzing the contribution of overweight/obesity and
GWG classification to the incidence of LGA infants (Fig. 1A),
SGA infants (Fig. 1B), caesarean sections (Fig. 1C) and
preeclampsia (Fig. 1D) by GDM status, the GWG and BMI
class were associated with LGA infants regardless of GDM
status (Fig. 1A, P < 0.0001) and with SGA infants only in
women with- out GDM (Fig. 1B, P < 0.01). Of note, the mean
rate of SGA was
13.7% in women without GDM whereas, when GWG was < 7
kg with normal weight, overweight and obesity, the rates
were
16.3%, 12.8% and 13.6%, respectively. There were no associations between GWG, BMI class, caesarean section (Fig.
1C) and preeclampsia (Fig. 1D) in women with and without
GDM.
3.4. Factors independently associated with LGA and
SGA
infants
The probability of delivering an LGA infant was associated with the following maternal and pregnancy
characteristics: positive association with increasing age
(29.7 5.9 years in women without an LGA infant vs 30.2
5.7 years in those with an LGA infant; P < 0.01), BMI
(24.6 4.7 kg/m2 vs
24.9 4.8 kg/m2 , respectively; P < 0.05), multiparity
(58.4% vs 73.2%, respectively; P < 0.0001), family history of
diabetes (20.4% vs 24.4%, respectively; P < 0.05), previous
pregnancy with macrosomia (2.1% vs 11.3%, respectively; P
< 0.0001), GDM (12.6% vs 23.3%, respectively; P < 0.0001)
and GWG classification (P < 0.001); negative association
with smoking before and during pregnancy (9.9% vs
5.6%, respectively; P < 0.0001), sub-Saharan African
(24.0% vs 19.8%, respec- tively; P < 0.0001) and Caribbean
(9.1% vs 6.6%, respectively; P < 0.0001) origins, and
preeclampsia (2.3% vs 11.0%, respec- tively; P < 0.01).
Multivariate analyses taking into account these parameters to
explain LGA infants showed that all of these factors, except

age, were independently associated with LGA infants,

including BMI and GWG (Table 2); this was also true when
only women without GDM were considered. Factors independently associated with LGA in women with GDM were
age, multiparity, previous pregnancy with macrosomia, and
GWG

11.516 kg and > 16 kg, but with none of the other parameters,
including BMI (Table 2).
The probability of delivering an SGA infant was associated
with BMI classification, multiparity (60.0% in women without an SGA infant vs 57.7% in those with; P < 0.05), smoking
before and during pregnancy (9.3% vs 10.7%, respectively;
P < 0.05), family history of diabetes (21.2% vs 17.8%, respectively; P < 0.0001), previous pregnancy with macrosomia
(3.1% vs 1.9%, respectively; P < 0.01), preeclampsia (1.9% vs
3.5%, respectively; P < 0.001) and GWG class (P < 0.01).
On mul- tivariate analyses, all of these parameters, except
multiparity, were associated with SGA infants, including BMI
(negative association) and GWG < 7 kg (positive association;
Table 3). On multivariate analyses performed according to
GDM status, overweight/obese and GWG classes remained
independently associated with SGA infants in women without
GDM, but no longer for women with GDM (Table 3).
4. Discussion
Our present findings confirm and extend previous reports
linking GDM, high maternal BMI and GWG with pregnancy
outcomes. In our large European, non-Asian cohort of women
without pregestational diabetes or hypertension, both pregravid
BMI and GWG were associated with LGA and SGA infants in
women without GDM. In contrast, in women with treated
GDM, overweight/obesity was not independently associated
with LGA and SGA infants, and GWG was only associated
with LGA infants. Considered altogether, these results suggest
that both overweight/obesity and GWG are crucial for fetal

growth in women without GDM, whereas GWG is the main

additional contributor in GDM to fetal overgrowth, with the
role of BMI blunted in women treated for GDM.
4.1. GDM prevalence, BMI excess and GWG in our
European cohort
As previously reported, GDM prevalence was high in our
followed-up women [2]. This was due to our diagnostic
criteria, which used low thresholds even before our adoption
of Inter- national Association of Diabetes and Pregnancy
Study Groups (IADPSG) criteria for defining GDM in 2011. In
addition, many of our patients have risk factors for GDM: 20%
of our women are from North Africa [21] and are particularly
vulnerable. Indeed, it was recently reported that more than half
the women with GDM in the four largest maternity units in our
area had psychosocial insecurity [24,25]. A large proportion of
women in our cohort began their pregnancies while
overweight (26.2%) or obese (13.9%); these are among the
highest rates in France and most likely due to the large
proportion of deprived populations living in our area. In
addition, the proportion of women of childbearing age with
obesity is increasing in France [26,27]; it was recently reported
that the proportion of pregnant women with overweight or
obesity had increased from 30.8% in 2002 to 37.6% in 2010 at
our centre [2]. The prevalence of obesity was higher than that
of FriuliVenezia Giulia, a region in northeastern Italy [9], and
almost the same as that reported by the international HAPO
study

E. Cosson et al. / Diabetes & Metabolism xxx (2015) xxxxxx

Fig. 1. Crude prevalences in women with and without gestational diabetes mellitus (GDM) of (A) large-for-gestational-age (LGA) and (B) small-for-gestational-age
(SGA) infants, (C) caesarean section and (D) preeclampsia, according to body mass index (obesity, overweight and normal weight) and gestational weight gain.

E. Cosson et al. / Diabetes & Metabolism xxx (2015) xxxxxx

Table 2
Parameters associated with large-for-gestational-age (LGA) infants in the total cohort and in women without and with gestational diabetes mellitus (GDM).
Total cohort

Women without GDM

Women with GDM

Multivariate analysis

Multivariate analysis

Multivariate analysis

OR [95 CI]

Age (10 years)

OR [95 CI]

NS
2

OR [95 CI]

NS

1.29 [1.071.51]

< 0.01

Maternal BMI (10 kg/m )

1.12 [1.001.24]

< 0.05

1.17 [1.031.31]

< 0.001

Multiparity (%)

1.87 [1.642.14]

< 0.001

1.90 [1.642.21]

< 0.001

0.48 [0.370.61]

REF
NS
< 0.001

0.42 [0.320.56]

REF
NS
< 0.001

REF
NS
NS

Maternal smoking
No (%)
Before (%)
Before and during (%)
Ethnicity
Caucasian (%)
Sub-Saharan African (%)
Caribbean (%)
Other (%)

NS
1.74 [1.302.35]

< 0.001

0.74 [0.640.89]
0.63 [0.500.80]

REF
< 0.001
< 0.001
NS

0.70 [0.590.83]
0.64 [0.490.83]

REF
< 0.001
< 0.001
NS

REF
NS
NS
NS

Family history of diabetes (%)

1.16 [1.011.33]

< 0.05

1.18 [1.011.38]

< 0.05

NS

Previous pregnancy with macrosomia (%)

4.52 [3.656.00]

< 0.001

4.86 [3.826.19]

< 0.001

3.78 [2.395.96]

< 0.001

GDM (%)

2.01 [1.752.32]

< 0.001

Preeclampsia (%)

0.49 [0.290.83]

< 0.01

0.42 [0.210.84]

< 0.05

NS

1.74 [1.492.03]
3.42 [2.834.13]

NS
REF
< 0.001
< 0.001

1.64 [1.381.95]
3.58 [2.914.40]

NS
REF
< 0.001
< 0.001

NS
REF
< 0.001
< 0.001

GWG classication
< 7 kg (%)
711.5 kg (%)
11.616 kg (%)
> 16 kg (%)

2.14 [1.542.97]
2.65 [1.684.17]

Mutivariate analyses used a logistic-regression model including the above factors associated with LGA infants and P < 0.10 on univariate analyses; NS: not
significant; REF: reference group.

Table 3
Parameters associated with small-for-gestational-age (SMA) infants in the total cohort, and in women without and with gestational diabetes mellitus (GDM).

BMI classication
Normal weight (%)
Overweight (%)
Obesity (%)
Multiparity (%)
Maternal smoking
No (%)
Before (%)
Before and during (%)

Total cohort

Women without GDM

Women with GDM

Multivariate analysis

Multivariate analysis

Multivariate analysis

OR [95 CI]

0.87 [0.780.97]
0.83 [0.720.96]

REF
< 0.05
< 0.05
NS

OR [95 CI]

OR [95 CI]

0.86 [0.770.98]
0.84 [0.720.98]
0.90 [0.821.00]

REF
< 0.05
< 0.05
0.05

REF
NS
NS
NS
REF
NS
NS

1.2 [1.031.4]

REF
NS
< 0.05

1.2 [1.031.4]

REF
NS
< 0.05

Family history of diabetes (%)

0.81 [0.720.91]

< 0.001

0.84 [0.740.95]

< 0.01

0.65 [0.470.91]

< 0.05

Preeclampsia (%)

1.87 [1.442.44]

< 0.001

1.90 [1.422.54]

< 0.001

1.85 [1.013.40]

0.05

1.13 [1.011.26]

< 0.05
REF
NS
NS

1.14 [1.011.29]

< 0.05
REF
NS
NS

GWG classication
< 7 kg (%)
711.5 kg (%)
11.616 kg (%)
> 16 kg (%)

NS
REF
NS
NS

Multivariate analyses used a logistic-regression model including the above factors, which were associated with SGA and P < 0.10 on univariate analyses, plus
previous pregnancy with macrosomia; REF: reference group; NS: not significant.

E. Cosson et al. / Diabetes & Metabolism xxx (2015) xxxxxx

[7], in which 13.7% had a BMI 33 kg/m2 at inclusion.

How- ever, obesity is still far less than reported in recent US
reports, which ranged from 20.0% [28] to 31.9% [29]. Indeed,
as in the US [15], our study shows that race/ethnicity is a
determinant of overweight/obesity.
As previously reported, women with overweight and obesity had less GWG than those with a normal BMI [5,8,13,14].
As women diagnosed with GDM are followed-up with dietary
counselling, this probably explains why this is so [30]. For this
reason, our study analyzed the incidence of pregnancy events
according to mutually exclusive BMI/GWG groups in women
without GDM and then in those treated for GDM.
4.2. Roles of BMI and GWG on fetal growth in women
with and without GDM
As in the five studies looking at the triumvirate of GDM,
BMI and GWG, multivariate analyses of our total cohort
showed that these three factors were independently associated
with LGA infants. Interestingly, Black et al. [5] recently
reported on the contributions of BMI, mild untreated GDM
and GWG to fetal overgrowth: 21.6% of LGA infants were
attributable to maternal overweight or obesity and 23.3% to
overweight or obesity with GDM, with an increasing GWG
associated with a greater preva- lence of LGA in all groups.
Similarly, Kim et al. [10] showed, in another US cohort, that
for all racial or ethnic groups, GDM contributed the least,
whereas GWG contributed the most, to LGA. This result was
partly due to a notably high prevalence of overweight/obesity
in that cohort. Indeed, their calculation of the partial
population attributable fraction took into account both the
prevalence and OR of a given risk factor, and was interpreted
as the proportion of cases that would be prevented if it were
pos- sible to eliminate that risk factor from the population
[10]. As the partial population attributable fraction of GDM,
BMI and GWG was < 50%, other factors may be important
contributors to LGA infants. Similar to the findings of others,
our data indi- cate that, while ethnicity [15,16], a familial
history of diabetes and personal history of a child with
macrosomia [15,29] suggest genetic causes of macrosomia,
smoking habits [6], multiparity and preeclampsia [31] are also
possible, non-genetic causes of macrosomia. All these factors
may play a role through epigenetic mechanisms, with fetal
epigenetic programming of adipokines involved when
considering BMI and GWG [32].
The effects of BMI and GWG may vary according to GDM
status. Our present study found that the effects of BMI and
GWG were greatest in patients without GDM, who received
no specific care in our maternity unit at the time. This has been
consistently reported in women without GDM [5,15,16,28].
For example, Di Benedetto et al. [8] showed, in an Italian
cohort, that the effect of overweight and obesity was present
only in glucose-tolerant women who had excess weight gain
during pregnancy. This suggests that the risk of macrosomia
associated with overweight/obesity might be limited by wellcontrolled GWG. The counterpart of a low GWG would be a
higher inci- dence of SGA infants [14,18]. In our cohort,
however, this effect was blunted in overweight/obese women,
who were unlikely to

deliver SGA infants. The increase in SGA infants associated

with GWG < 7 kg was balanced by a decreased incidence of
overweight/obese women. To illustrate this point, GWG < 7
kg was associated with the highest prevalence of SGA infants
only in lean women whereas, in the overweight/obese women,
SGA rates were similar to the mean rate of the overall cohort.
These findings suggest that closer monitoring of GWG in
women with pregravid BMIs 25 kg/m2 may be warranted
to prevent LGA infants with no negative impact on SGA
births, as rec- ommended by the IOM [18]. Nevertheless,
meta-analyses also show that, although antenatal interventions
for overweight or obesity can limit GWG [33], the outcomes
are often unchanged [33,34].
Our present results differed in women with GDM, who
were followed-up to control both their diets and blood glucose
levels. In fact, as found by others [30], these women had less
GWG than women without GDM. Also, in these women, BMI
no longer had any effect on LGA births. Indeed, the effect of
BMI might be driven by GDM, which is more frequent when
BMI is increased [3,15]. Otherwise, GWG remained an
important contributor to fetal overgrowth in this
subpopulation, with the same result as found in an Italian
cohort. When only women with GDM were considered, GWG,
but neither overweight nor obesity, were associated with
macrosomia [9]. However, an inde- pendent effect of GWG
and obesity on LGA infants was found in three American
studies [16,28,29]. This was also the case of the study by
Black et al. [5] although, in that study, GDM was treated by
neither diet nor medication.
4.3. Other outcomes
As in previous reports, GDM [1,5,9,10,15] and obesity
[4,5,7,8] were associated in our cohort with more preeclampsia, and GDM with more caesarean section, after examining
the contributions of BMI and GWG to these outcomes
according to GDM status. Regarding preeclampsia, there was
no influence of overweight/obesity and GWG classification,
thus stressing the role of GDM per se. Also, no association
was identified between BMI and GWG classes for caesarean
section regard- less of GDM status, whereas such an
association has often been reported [4,5,7,8,12]. Several
population-based studies found that planned and especially
emergency caesarean delivery was significantly increased with
increasing BMI [35]. However, these studies did not stratify
patients according to GDM status. Caregivers were inclined to
perform caesarean sections in obese patients because of
concerns about obesity-related macrosomia, and perinatal
complications such as shoulder dystocia [5] and perinatal
mortality [36].
4.4. Strengths and weaknesses of the
study
Our study involved a large sample size, thereby giving
power to its significant differences, and allowing multivariate
analyses and strict selection criteria, including no known
hypertensive disorders or diabetes before pregnancy, no
underweight women, no women of Asian origin and no
twin pregnancies. This removed the major sources of

potential confounders. The data

E. Cosson et al. / Diabetes & Metabolism xxx (2015) xxxxxx

came from a single institution with a comprehensive and

consis- tent perinatal care programme, and the multiethnic
population included a large proportion of women living with
psychoso- cial vulnerability [24,25]. All information was
collected at a university hospital, precluding generalization of
the results.
Regarding the parameters of interest, although GDM
screen- ing was universal, some women were not screened.
Therefore, our study considered the presence of GDM in the
intention- to-screen population, as done in other studies
[10,15]. The proportion of unscreened women was 12.5% in
2011 at our hospital and is likely to have remained stable over
the past decade. GDM was not defined according to IADPSG
criteria. However, the association between BMI and
pregnancy out- comes is reportedly not influenced by the
definition of GDM [37,38], and our cohort had a GDM
prevalence rate similar to that of the IADPSG criteria. Our
results were not adjusted for glucose control, as these data
were not available. BMI was cal- culated from measured
height, but weight before pregnancy was self-reported.
One strength of our study is that GWG was not considered
according to BMI, as that would have led to consideration of
the BMI effect twicefirst per se and then through excessive
GWG, defined according to normal weight, overweight and
obesity sta- tus. However, GWG was not adjusted for
gestational age at delivery, which is difficult as GWG is not
linear during preg- nancy, but increases during the lattermost
weeks of pregnancy. Finally, LGA and SGA infants were
defined by comparison to the general French population [22]
with similar thresholds across ethnicities. However, their
determinants were adjusted for ethnicity.
5. Conclusion
In the context of the current escalation of obesity [26,27]
and GDM [39], our present study confirms that GDM, even
when treated, is associated with adverse pregnancy outcomes.
In our European cohort of women with GDM, GWG is
additionally and independently associated with more LGA
infants, whereas overweight/obesity is not. This suggests that,
even after rein- forcing GWG control in women treated for
GDM, the most this would achieve does not appear to include
more SGA infants in this pregnant population. However,
weight loss subsequent to a diagnosis of GDM has recently
been reported to be associated with a 1.69-fold increased rate
of SGA infants [40].
In women without GDM, our data show that considering
pre- gravid BMI and GWG is crucial for estimating the risk for
LGA infants. Our observational results suggest that dietary
advice could be offered to overweight/obese women before
pregnancy to reduce the risk of preeclampsia and LGA as well
as the risk of GDM. In women without GDM, controlling
weight gain during pregnancy might limit the incidence of
LGA without inducing fetal restriction in those who are
overweight/obese. It was recently shown that a low-intensity
lifestyle intervention in women at high risk for GDM
optimalized healthy GWG, while limiting weight gain was
more effective in overweight women [41].

Disclosure of interest
The authors declare that they have no conflicts of interest
concerning this article.

Acknowledgements
We thank Prof Eric Vicaut (APHP, Unit of Clinical
Research, Lariboisire Hospital, Paris 7 University, Paris,
France) for his help with the statistical analyses.

References
[1] World Health Organization. Diagnostic criteria and classification of
hyperglycaemia first detected in Pregnancy; 2013 http://apps.who.
int/iris/bitstream/10665/85975/1/WHO NMH MND 13.2 eng.pdf
[2] Cosson E, Benbara A, Pharisien I, Nguyen MT, Revaux A, Lormeau B, et
al.
Diagnostic and prognostic performances over 9 years of a selective
screen- ing strategy for gestational diabetes mellitus in a cohort of 18,775
subjects. Diabetes Care 2013;36:598603.
[3] Galtier F. Definition, epidemiology, risk factors. Diabetes Metab
2010;36:62851.
[4] Hyperglycaemia and adverse pregnancy outcome (HAPO) study: associations with neonatal anthropometrics. Diabetes 2009;58:4539.
[5] Black MH, Sacks DA, Xiang AH, Lawrence JM. The relative contribution of prepregnancy overweight and obesity, gestational weight gain,
and IADPSG-defined gestational diabetes mellitus to fetal overgrowth.
Dia- betes Care 2013;36:5662.
[6] Djelantik AA, Kunst AE, van der Wal MF, Smit HA, Vrijkotte TG.
Contri- bution of overweight and obesity to the occurrence of adverse
pregnancy outcomes in a multi-ethnic cohort: population attributive
fractions for Ams- terdam. BJOG 2012;119:28390.
[7] Catalano PM, McIntyre HD, Cruickshank JK, McCance DR, Dyer AR,
Metzger BE, et al. The hyperglycaemia and adverse pregnancy outcome
study: associations of GDM and obesity with pregnancy outcomes. Diabetes Care 2012;35:7806.
[8] Di Benedetto A, DAnna R, Cannata ML, Giordano D, Interdonato ML,
Corrado F. Effects of prepregnancy body mass index and weight gain
dur- ing pregnancy on perinatal outcome in glucose-tolerant women.
Diabetes Metab 2012;38:637.
[9] Alberico S, Montico M, Barresi V, Monasta L, Businelli C, Soini V, et al.
The role of gestational diabetes, pre-pregnancy body mass index and gestational weight gain on the risk of newborn macrosomia: results from a
prospective multicentre study. BMC Pregnancy Childbirth 2014;14:23.
[10] Kim SY, Sharma AJ, Sappenfield W, Wilson HG, Salihu HM. Association of maternal body mass index, excessive weight gain, and gestational
diabetes mellitus with large-for-gestational-age births. Obstet Gynecol
2014;123:73744.
[11] Oteng-Ntim E, Kopeika J, Seed P, Wandiembe S, Doyle P. Impact of
obesity on pregnancy outcome in different ethnic groups: calculating
population attributable fractions. PLoS One 2013;8:e53749.
[12] Gilead R, Yaniv Salem S, Sergienko R, Sheiner E. Maternal isolated obesity and obstetric complications. J Matern Fetal Neonatal Med
2012;25:257982.
[13] Blomberg M. Maternal and neonatal outcomes among obese women with
weight gain below the new Institute of Medicine recommendations.
Obstet Gynecol 2011;117:106570.
[14] Hinkle SN, Sharma AJ, Dietz PM. Gestational weight gain in obese
mothers and associations with fetal growth. Am J Clin Nutr
2010;92:64451.
[15] Bowers K, Laughon SK, Kiely M, Brite J, Chen Z, Zhang C. Gestational diabetes, pre-pregnancy obesity and pregnancy weight gain in
relation to excess fetal growth: variations by race/ethnicity. Diabetologia
2013;56:126371.

E. Cosson et al. / Diabetes & Metabolism xxx (2015)

xxxxxx
9
[16] Ellerbe CN, Gebregziabher M, Korte JE, Mauldin J, Hunt KJ.
Quantifying the impact of gestational diabetes mellitus, maternal weight
and race on birthweight via quantile regression. PLoS One
2013;8:e65017.
[17] Alberti KG, Eckel RH, Grundy SM, Zimmet PZ, Cleeman JI, Donato
KA, et al. Harmonizing the metabolic syndrome: a joint interim
statement of the International Diabetes Federation Task Force on
Epidemiology and Prevention; National Heart, Lung, and Blood Institute;
American Heart Association; World Heart Federation; International
Atherosclerosis Soci- ety; and International Association for the Study
of Obesity. Circulation
2009;120:16405.
[18] Rasmussen KM, Catalano PM, Yaktine AL. New guidelines for weight
gain during pregnancy: what obstetrician/gynecologists should know.
Curr Opin Obstet Gynecol 2009;21:5216.
[19] Langer O, Yogev Y, Xenakis EM, Brustman L. Overweight and obese
in gestational diabetes: the impact on pregnancy outcome. Am J Obstet
Gynecol 2005;192:176876.
[20] Cosson E, Benchimol M, Carbillon L, Pharisien I, Paries J, Valensi P, et
al.
Universal rather than selective screening for gestational diabetes mellitus
may improve fetal outcomes. Diabetes Metab 2006;32:1406.
[21] Cosson E, Cussac-Pillegand C, Benbara A, Pharisien I, Jaber Y, Banu I,
et al. The diagnostic and prognostic performance of a selective screening
strategy for gestational diabetes mellitus according to ethnicity in
Europe. J Clin Endocrinol Metab 2014;99:9961005.
[22] Leroy B, Lefort F. [The weight and size of newborn infants at birth]. Rev
Fr Gynecol Obstet 1971;66:3916.
[23] Spong CY, Beall M, Rodrigues D, Ross MG. An objective definition of
shoulder dystocia: prolonged head-to-body delivery intervals and/or the
use of ancillary obstetric maneuvers. Obstet Gynecol 1995;86:4336.
[24] Bihan H, Cosson E, Khiter C, Vittaz L, Faghfouri F, Leboeuf D, et al.
Fac- tors associated with screening for glucose abnormalities after
gestational diabetes mellitus: baseline cohort of the interventional
IMPACT study. Diabetes Metab 2014;40:1517.
[25] Cosson E, Bihan H, Vittaz L, Khiter C, Carbillon L, Faghfouri F, et al.
Improving postpartum glucose screening after gestational diabetes mellitus: a cohort study to evaluate the multicentre IMPACT initiative. Diabet
Med 2015;32:18997.
[26] Diouf I, Charles MA, Ducimetiere P, Basdevant A, Eschwege E, Heude
B.
Evolution of obesity prevalence in France: an age-period-cohort analysis.
Epidemiology 2010;21:3605.
[27] Eschwege E, Basdevant A, Crine A, Moisan C, Charles MA. Type 2
diabetes mellitus in France in 2012: results from the ObEpi survey.
Diabetes Metab
2015;41:5561.
[28] Kim C. Maternal outcomes and follow-up after gestational diabetes
melli- tus. Diabet Med 2014;31:292301.

[29] Ouzounian JG, Hernandez GD, Korst LM, Montoro MM, Battista LR,
Walden CL, et al. Pre-pregnancy weight and excess weight gain are risk
factors for macrosomia in women with gestational diabetes. J Perinatol
2011;31:71721.
[30] Bo S, Menato G, Bardelli C, Lezo A, Signorile A, Repetti E, et al. Low
socioeconomic status as a risk factor for gestational diabetes. Diabetes
Metab 2002;28:13940.
[31] Eskild A, Romundstad PR, Vatten LJ. Placental weight and birthweight: does the association differ between pregnancies with and without
preeclampsia? Am J Obstet Gynecol 2009;201:595e15.
[32] Houde AA, Hivert MF, Bouchard L. Fetal epigenetic programming of
adipokines. Adipocyte 2013;2:416.
[33] Quinlivan JA, Julania S, Lam L. Antenatal dietary interventions in
obese pregnant women to restrict gestational weight gain to Institute of
Medicine recommendations: a meta-analysis. Obstet Gynecol 2011;118:
1395401.
[34] Dodd JM, Grivell RM, Crowther CA, Robinson JS. Antenatal interventions for overweight or obese pregnant women: a systematic review of
randomised trials. BJOG 2010;117:131626.
[35] Ovesen P, Rasmussen S, Kesmodel U. Effect of prepregnancy maternal overweight and obesity on pregnancy outcome. Obstet Gynecol
2011;118:30512.
[36] Cresswell JA, Campbell OM, De Silva MJ, Filippi V. Effect of maternal
obesity on neonatal death in sub-Saharan Africa: multivariable analysis
of
27 national datasets. Lancet 2012;380:132530.
[37] Owens LA, OSullivan EP, Kirwan B, Avalos G, Gaffney G, Dunne F.
ATLANTIC DIP: the impact of obesity on pregnancy outcome in
glucose- tolerant women. Diabetes Care 2010;33:5779.
[38] Dennedy MC, Avalos G, OReilly MW, OSullivan EP, Gaffney G, Dunne
F.
ATLANTIC-DIP: raised maternal body mass index (BMI) adversely
affects maternal and fetal outcomes in glucose-tolerant women according
to Inter- national Association of Diabetes and Pregnancy Study Groups
(IADPSG) criteria. J Clin Endocrinol Metab 2012;97:E60812.
[39] Feig DS, Hwee J, Shah BR, Booth GL, Bierman AS, Lipscombe LL.
Trends in incidence of diabetes in pregnancy and serious perinatal
outcomes: a large, population-based study in Ontario, Canada, 1996
2010. Diabetes Care 2014;20:17683.
[40] Yee LM, Cheng YW, Inturrisi M, Caughey AB. Gestational weight loss
and perinatal outcomes in overweight and obese women subsequent
to diagnosis of gestational diabetes mellitus. Obesity (Silver Spring)
2013;21:E7704.
[41] Harrison CL, Lombard CB, Strauss BJ, Teede HJ. Optimizing healthy
gestational weight gain in women at high risk of gestational diabetes: a
randomized controlled trial. Obesity (Silver Spring) 2013;21:9049.