See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/224932248

Prevalence and risk factors of bacterial

vaginosis during the first trimester of
pregnancy in a large French population-based
study

Article in European journal of obstetrics, gynecology, and reproductive biology May 2012
DOI: 10.1016/j.ejogrb.2012.04.007 Source: PubMed

CITATIONS READS

14 159

7 authors, including:

David Desseauve Babak Khoshnood

University Hospital of Lausanne French Institute of Health and Medical Resea
24 PUBLICATIONS 32 CITATIONS 222 PUBLICATIONS 3,748 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Rothschild-Besanon Cohort View project

Epidemiology of Congenital heart and Neural tube defects View project

All content following this page was uploaded by David Desseauve on 26 December 2014.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
G Model
EURO-7689; No. of Pages 5

European Journal of Obstetrics & Gynecology and Reproductive Biology xxx (2012) xxxxxx

Contents lists available at SciVerse ScienceDirect

European Journal of Obstetrics & Gynecology and

Reproductive Biology
journal homepage: www.elsevier.com/locate/ejogrb

Prevalence and risk factors of bacterial vaginosis during the first trimester of
pregnancy in a large French population-based study
D. Desseauve a,b,c,*, J. Chantrel c,e, A. Fruchart d,e, B. Khoshnood a,b,e, G. Brabant f,e,
P.Y. Ancel a,b,e, D. Subtil a,b,c,e
a
INSERM, UMR S953, IFR 69, Unite de Recherche Epidemiologique en Sante Perinatale et Sante des Femmes et des Enfants, Hopital Tenon, F-75020 Paris, France
b
UPMC Univ Paris 06, UMR S 953, F-75005 Paris, France
c
Hopital Jeanne de Flandre, Universite Lille II, 1 rue Eugene Avinee, 59037 Lille Cedex, France
d
Laboratoire de Bacteriologie-Hygiene, Universite Lille 2, 1 rue Eugene Avinee, 59037 Lille, France
e
Laboratoire de Bacteriologie, Centre Hospitalier, avenue Desandrouin, 59300 Valenciennes, France
f
Hopital Saint Vincent, GHICL, 59046 Lille, France

A R T I C L E I N F O A B S T R A C T

Article history: Objectives: Bacterial vaginosis is a risk factor for preterm delivery. Its prevalence and risk factors in
Received 14 May 2011 Europe are not well known. Our objective was to assess both in early pregnancy.
Received in revised form 14 December 2011 Study design: As part of the PREMEVA randomized controlled trial, this population-based study included
Accepted 5 April 2012
14,193 women screened before 14 weeks gestation for bacterial vaginosis in the 160 laboratories of the
Nord-Pas-de-Calais region in France. Bacterial vaginosis was defined by a Nugent score ! 7. Data were
Keywords: collected about maternal tobacco use, age, education, and history of preterm birth. We estimated the
Population-based study
prevalence of bacterial vaginosis and used a multilevel logistic regression model to identify significant
Prevalence
risk factors for it.
Risk factor
Bacterial vaginosis Results: Among the 14,193 women assessed before 14 weeks gestation, the prevalence of bacterial
vaginosis was 7.1% (95% CI: 6.67.5%). In the multivariate analysis, smoking during pregnancy tobacco
(adjusted OR: 1.38; 95% CI: 1.191.60), maternal age 1819 years (adjusted OR: 1.40; 95% CI: 1.011.93),
and educational level (completed only primary school: adjusted OR: 1.77; 95% CI: 1.352.31; completed
only secondary school: adjusted OR: 1.27; 95% CI: 1.101.48) were independent risk factors for bacterial
vaginosis. History of preterm delivery was not an independent risk factor of bacterial vaginosis: adjusted
OR: 1.15; 95% CI: 0.901.47.
Conclusion: In a large sample of women in their first trimester of pregnancy in France, the prevalence of
bacterial vaginosis was lower than rates reported in other countries, but risk factors were similar: young
age, low level of education, and tobacco use during pregnancy. These results should be considered in
future strategies to reduce preterm delivery.
! 2012 Elsevier Ireland Ltd. All rights reserved.

1. Introduction flora is discovered, the stronger the association between BV and

Each year, 1 billion women have a urinary or lower genital tract
infection, most frequently bacterial vaginosis (BV) [1], a modifica-
tion of the vaginal flora that is associated with an increased risk of
preterm delivery and spontaneous miscarriage during the first and
second trimesters [24]. The nature of this association has not been
clearly elucidated, but many authors suggest that preterm delivery
results from bacteria ascending from the vagina to the membranes
and amniotic fluid [5]. The earlier in pregnancy that this abnormal
* Corresponding author at: INSERM U953 Batiment Recherche Hopital Tenon, 4,
Rue de la Chine, 75020 Paris, France. Tel.: +33 01 42 34 55 70;
fax: +33 01 43 26 89 79.
E-mail address: desseauve.d@gmail.com (D. Desseauve).
preterm delivery appears to be [6]. Ascension of vaginal micro-
organisms may thus occur as early as the first trimester. This
implies that pregnant women should be tested for vaginosis as
early as possible to prevent second trimester miscarriage or
preterm delivery [7]. Recent investigations report a potential
genetic susceptibility to lower genital tract infections [8].
The frequency of BV in low-risk pregnant women, i.e., those
with no history of preterm delivery, aged between 25 and 29 years,
is not well known. Studies in the US estimate the frequency of BV at
1639% [9,3,1014]. The reported frequency appears lower in
Europe, ranging from 1.6% to 28% [1528]. Most of these studies,
however, were conducted in highly selected samples, e.g., patients
undergoing in vitro fertilization or with threatened preterm
delivery, or they used non-standard tests to diagnose BV, i.e.,
Amsels clinical criteria [9,24] or Pap smears [18].

0301-2115/$ see front matter ! 2012 Elsevier Ireland Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ejogrb.2012.04.007

Please cite this article in press as: Desseauve D, et al. Prevalence and risk factors of bacterial vaginosis during the first trimester of
pregnancy in a large French population-based study. Eur J Obstet Gynecol (2012), http://dx.doi.org/10.1016/j.ejogrb.2012.04.007
G Model
EURO-7689; No. of Pages 5
2 D. Desseauve et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology xxx (2012) xxxxxx
Recent studies of risk factors for BV have focused mainly on
North American women at risk of preterm delivery. Reported risk
factors include low socioeconomic status, African descent, and
cigarette smoking [29,30]. Very few such studies have been
conducted in Europe. Because BV is one of the most serious
potential explanations for spontaneous pregnancy loss [7], we
sought to estimate its prevalence and to study its risk factors in a
large general population sample of women in their first trimester
of pregnancy in France.
2. Materials and methods
2.1. Population
This prospective study was the first part of the PREMEVA
project (Prevention of Very PREterM Delivery by Testing for and
Treatment of Bacterial VAginosis), a randomized controlled trial to
test the efficacy of clindamycin compared with placebo for
reducing the rate of births before 32 weeks gestation among
women with BV diagnosed early in pregnancy (NCT00642980.
2008). This study took place in all 160 medical laboratories in a
French region (Nord-Pas-de-Calais) of 4 million inhabitants and
approximately 58,000 births a year.
Each laboratory asked all pregnant women coming in for any
testing if they would be willing to take a self-sample to test their
vaginal microbiota. The inclusion criteria were maternal age of 18
years or older, ability to understand French, a gestational age at the
time of screening of less than 13 completed weeks of amenorrhoea,
and planned delivery in the Nord-Pas-de-Calais region. After an
oral explanation, each patient received an informational document
and provided written consent.
From April 2006 to August 2008, a total of 16,188 women were
screened for BV in the 160 participating laboratories. This
screening was conducted in association with obligatory examina-
tions performed during pregnancy in France: 1040 patients were
excluded because they did not meet one or more of the inclusion
criteria, and 210 women refused screening, and thus 14,938
women were eligible for the study. Among them, 713 had an
uninterpretable examination and 32 had missing Nugent scores.
The analysis finally included 14,193 women.
2.2. Main outcome measure
The main outcome measure was BV, diagnosed according to the
Nugent score. Each patient received a swab to take a vaginal self-
sample, as well as a diagram and instructions that explained in
detail how to use it: she was to introduce it gently into the vagina,
as she would a tampon, to a depth of approximately 5 cm, remove
it, place it in its tube, close the tube, and give it to the clinical
pathologist. The women took these self-samples at the laboratory.
The vaginal secretions were spread on a clean slide and heat-fixed
within 4 h. The vaginal smear was then gram-stained. Bacterial and
cellular abundance was assessed by examining several microscop-
ic fields at a 10" magnification. On 10 fields with abundant
microorganisms, three bacterial morphotypes were identified and
quantified by the number of bacteria visible at a 100" magnifica-
tion: lactobacillus (Gram-positive bacilli with regular parallel
edges), Gardnerella and other anaerobic morphotypes (Gram-
variable polymorphous bacilli) and morphotype Mobiluncus
(Gram-variable rod-shaped bacilli). The sum of the scores obtained
for each morphotype constitutes the NKH (NugentKrohnHillier)
score [31].
This method has been used successfully in other studies [25].
Moreover, a preliminary study in our region showed the good
quality of the scores and the acceptability of this procedure to
pregnant women [32]. In this pilot study, the quality and
Please cite this article in press as: Desseauve D, et al. Prevalence and risk factors of bacterial vaginosis during the first trimester of
pregnancy in a large French population-based study. Eur J Obstet Gynecol (2012), http://dx.doi.org/10.1016/j.ejogrb.2012.04.007
interpretability of the samples was acceptable in 93% of 339
samples taken. The Nugent score was chosen because it is a
sensitive diagnostic method that can be used routinely [31].
Because this score was not commonly used in France, onsite
training for all of the medical laboratories in the region was a
prerequisite to this study.
Scores from 0 to 6 corresponded to normal or intermediate
flora, and scores from 7 to 10 were classified as BV [31]. The
presence of clue cells was very suggestive, but neither necessary
nor sufficient for a BV diagnosis. The score could not be calculated
(n = 713) when the bacterial density was insufficient or when there
was a proliferation of a single microbe of a morphotype not
included in the score, as, for example, in the case of contamination.
Two separate steps were taken to ensure the quality and verify
the correct application of the screening method chosen. First, the
laboratories underwent onsite training and were evaluated at the
conclusion of the training. All private and public medical
laboratories in the Nord-Pas-de-Calais region were trained to
read and interpret the NKH score. Between January and November
2006, 160 laboratories underwent personalized onsite training and
supervised practice by a technician and a microbiologist from the
Lille University Hospital Center; they also received a training video,
a technical brochure, and collection slides. Second, the quality
control programme consisted of a rereading of the first 2870 slides
by an expert clinical pathologist selected by the PREMEVA steering
committee member responsible for laboratory training. The
resulting Nugent scores were compared with those of the
laboratories by calculating Cohens Kappa coefficient, to measure
their concordance. Concordance was very good, with a Kappa
coefficient of 0.89, 95% CI = [0.850.92].
The patients with BV were then invited by their general
practitioner to participate in the randomized trial [33]. The
regional ethics committee approved the protocol (Session of 5
October 2004).
2.3. Factors studied
Women who agreed to participate in the first stage of the study
(testing for BV) were asked by a laboratory employee for their age,
educational level (primary, secondary, or post-secondary), wheth-
er they had smoked since the beginning of pregnancy, and any
history of preterm delivery. A short standardized questionnaire
was used to collect the data.
2.4. Statistical analysis
We first estimated the prevalence of BV and then studied its
variations according to the patients characteristics, i.e., mater-
nal age, educational level, obstetric history (previous preterm
birth), and tobacco use. Crude associations between each
individual factor and BV were quantified by odds ratios (ORs)
and their 95% confidence intervals (CIs). Multivariate analysis
followed to allow us to obtain a better estimate of the role of
each factor. Before modelling, we checked for the absence of
interactions between the explanatory variables. All variables
were included in the model. Because information on previous
preterm birth was missing for 1028 patients (7.1%), a class that
included missing values was generated to limit the loss of
information on the other variables in the multivariate model.
Two categories of variables were considered: those characteriz-
ing the women (individual variables: maternal age, tobacco
consumption, educational level, history of preterm delivery) and
the variables characterizing the laboratories (identification
number and their location in one of four zones: Artois, Hainaut,
Coast area, and Lille metropolitan area, each with a somewhat
different distribution of socioeconomic characteristics).
G Model
EURO-7689; No. of Pages 5

D. Desseauve et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology xxx (2012) xxxxxx 3

Table 1 4. Comments
Characteristics of women included in the PREMEVA study and prevalence of
bacterial vaginosis.
To our knowledge, this study is the first large population-based
PREMEVA study (20062008) study in France to estimate the prevalence of BV and identify its
Maternal age (years) N = 14,193 risk factors early in pregnancy. We found that the prevalence of BV
Mean (sd) 28.5 (5.1) before 13 weeks was 7% and that the main risk factors were young
<20 3.5% maternal age, smoking during pregnancy, and low educational
2024 22.5%
level.
2529 37.3%
3034 25.1% The screening test was addressed to all pregnant women in
3539 9.7% their first trimester in the Nord-Pas-de-Calais area from 2006 to
!40 1.7% 2008. In all, 745 (4%) women were excluded from the sample
Tobacco consumption N = 14,034
because their Nugent score could not be estimated. The main
Yes 23.4%
Educational level N = 13,744
reasons were insufficient bacterial density, absence of lactobacil-
Primary 4.8% lus, a dried and unusable sample (because the swab was not
History of preterm delivery N = 13,165 introduced or the tube closed incorrectly), or a broken slide. The
Yes 7.9% finding that the women who were excluded were older and had a
sd = standard deviation. previous preterm delivery more frequently than the women
included in the analysis suggests that the excluded sample was at
higher risk than the women finally included (see the Table in
A mixed hierarchical model was used to take the hierarchical Appendix A).
structure of these data into account in the estimation of the The diagnoses of bacterial vaginosis from the laboratories were
parameters [34]. The data were analysed with STATA version 10. compared with those of expert microbiologists, and their concor-
dance was very good. The stability of the result was further verified
by calculating the Kappa coefficient in the groups of patients with
3. Results known obstetric risks (women <20 years or >35 years and
smokers).
Table 1 presents the characteristics of the women included in Finally the association between bacterial vaginosis and the
the PREMEVA study. Of 14,193 women in our study; 1003 women identified risk factors in the total sample (N = 14,193) was similar
had a Nugent score ! 7, for a BV prevalence of 7.1%, 95% CI [6.6 to that in the sample reread by the experts (N = 2720).
7.5]. In France, as in many other countries, data on the prevalence of
The prevalence of BV was higher among smokers (9.6%), women BV among low-risk pregnant women were previously unavailable.
aged 1819 years (11.0%) or 2024 years (8.9%), and women with Our results demonstrated a prevalence clearly lower than those
only a primary (11.5%) or secondary educational level (8.4%), reported by most other studies in Europe. Many of those results,
compared with non-smokers (6.3%), women aged 2529 years however, were not comparable because of the different techniques
(6.5%), and those with post-secondary education (5.8%) (Table 2). used to assess BV [1820,26]. Nonetheless, our results are
After adjustment, smoking during pregnancy (adjusted odds ratio, consistent with two recent European studies that did use the
aOR = 1.38 95% CI = [1.191.60]), maternal age 1819 years Nugent score and found BV prevalences of 4.5% [17] and 10% [25].
(aOR = 1.39 95% CI = [1.011.93]), and primary (aOR = 1.69 95% Both studies were conducted in low-risk women, i.e., populations
CI = [1.232.23]) and secondary (aOR = 1.26 95% CI = [1.081.46]) fairly comparable to ours. The prevalence of BV in the US studies
levels of education remained significantly associated with BV varied from 16% to 39% [3,1014]. Differences in the design of these
(Table 2). studies and in the characteristics of populations included might

Table 2
Factors associated with bacterial vaginosis: a multilevel analysis.

% bacterial vaginosis OR 95% CI p value Multilevel logistic regression,

N = 13,651 aOR 95% CI

Level 1: individual characteristics

Maternal age (years) (N = 14,193)
<20 11.0 1.80 [1.302.38] 0.02 1.39 [1.011.93]
2024 8.9 1.40 [1.181.64] 1.19 [1.001.42]
2529 6.5 1.00 1
3034 5.9 0.90 [0.751.08] 0.89 [0.741.07]
3539 6.6 1.00 [0.791.28] 0.97 [0.761.24]
!40 years 5.8 0.90 [0.501.51] 0.82 [0.471.43]
Smoked during pregnancy (N = 14,034)
No 6.3 1.00 <0.001 1
Yes 9.6 1.60 [1.391.84] 1.38 [1.191.61]
Educational level (N = 13,744)
Post-secondary 5.8 1.00 <0.0001 1
Completed secondary 8.4 1.50 [1.291.70] 1.26 [1.081.46]
Completed primary 11.5 2.12 [1.632.73] 1.69 [1.232.23]
History of preterm delivery (N = 13,165)
No 6.9 1.00 0.44 1
Yes 8.4 1.23 [0.981.55] 1.15 [0.901.47]
Missing values 7.1 1.02 [0.791.31] 0.96 [0.731.26]

The odds ratios are adjusted for the individual variables (maternal age, smoking during pregnancy, educational level, history of preterm delivery) and the variable of centre
(zone).

Please cite this article in press as: Desseauve D, et al. Prevalence and risk factors of bacterial vaginosis during the first trimester of
pregnancy in a large French population-based study. Eur J Obstet Gynecol (2012), http://dx.doi.org/10.1016/j.ejogrb.2012.04.007
G Model
EURO-7689; No. of Pages 5

4 D. Desseauve et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology xxx (2012) xxxxxx

explain the differences from European studies. In particular, the US screening for bacterial vaginosis [42]. In France, however,
studies included mainly women who were black or had a low screening and a treatment of bacterial vaginosis has been
educational level, both factors associated with BV [35]. recommended for women with previous adverse obstetric history
Moreover, the prevalence of BV varies according to the by the National Agency for Accreditation and Evaluation in Health
diagnostic method used. Amsels clinical criteria have a lower (ANAES). In all cases, the benefit of early treatment among low-risk
sensitivity than other diagnostic methods and may underestimate women is unknown; the PREMEVA study was conducted
the BV rate [13]. In a study of 492 asymptomatic pregnant women specifically to answer this question.
at low risk, Gratacos et al. reported a BV rate of 1.6% according to
Amsels criteria but a 4.5% rate according to the Nugent score [17]. 5. Conclusion
Spiegels score has the disadvantage of depending on numerous
bacterial morphotypes, including Gram-positive cocci [31]. In our Overall, little is known about the frequency of BV during the
study, BV was diagnosed with the Nugent score from self-sampled first trimester of pregnancy, and no study published in Europe or
swabs and all the laboratories received identical onsite training. the US of women before 15 weeks has previously used the Nugent
This score has the advantage of being easy, fast, reproducible, and score. By bringing together several thousand women, our study
inexpensive [31,32,36]. It is commonly used in studies of vaginosis made it possible to show that the frequency of BV at the beginning
and makes it possible to grade the disruption of vaginal flora. of pregnancy is lower than the estimates reached in studies of
In our study, patients with a history of preterm delivery had no other low-risk populations, which were often small or hospital-
increased risk of BV. We cannot rule out the possibility of based samples, or both [1820,26].
misclassification in reporting obstetric history. Because the This study also shows a significant association with a low
diagnosis of BV was not known when the questionnaire was educational level and smoking. These associations with preterm
completed, however, these errors are likely to be non-differential. birth were already known, but their relation to imbalanced vaginal
We found that the BV rate was significantly higher in patients who ecology was much less known. While it is difficult to intervene
smoked. This result is consistent with two Danish studies that rapidly and effectively on socioeconomic level, prevention
reported a risk of BV among smokers 5070% higher than in non- campaigns can be envisioned to reduce smoking at the beginning
smokers [23,24]. Two other studies have reported that this of pregnancy.
association was not statistically significant [19,37]. The mechanisms Because BV is considered a risk factor for very preterm birth, it is
that might link tobacco consumption and BV remain unclear, but useful to know its frequency and risk factors in order to apply
Pavlova and Tao [38] have suggested that the presence of preventive strategies from the first trimester.
benzopyrene diol epoxide in the vaginal secretions of smokers
significantly increases phagocytosis of lactobacilli, reducing their Acknowledgements
numbers and thus promoting the development of anaerobic bacteria.
We observed that the risk of BV was higher in women of lower The authors thank C. Nolf, J.C. Dugimont and the entire team
educational level. Other authors have found similar results, although from the North Picardy Regional Association of Clinical Patholo-
not to a significant extent [19,24]. Similarly, others have observed the gists (ABRNP) for their contribution to the success of the study.
relation we found between maternal youth and BV prevalence [19]. In They also thank C. Leignel, S. Deghilage and M.C. Bissinger, the
our study, the risk of BV increased as maternal age decreased. clinical research assistants for the study, as well as all of the clinical
Although pregnancy in young patients is often a marker of social pathologists and clinicians for their essential contribution to the
deprivation [39], young maternal age remained significantly success of this project.
associated with BV after adjustment for educational level. This result
could reflect insufficient adjustment for socioeconomic status; that is,
Appendix A
educational status alone may be an insufficient proxy for socioeco-
nomic status, or a maternal age effect may exist with a pathophysio-
logical mechanism that has not yet been elucidated. Comparison between women whose Nugent scores could and
This study has shown that approximately 7% of expectant could not be calculated.
mothers in France have vaginosis during their first trimester of Eligible Women without p
pregnancy. BV is associated with a significantly increased risk of women a Nugent score
preterm delivery [40], and this association increases as gestational
N 14,193 745
age at diagnosis decreases, i.e., the OR before 20 weeks is 4.2 (95% Maternal age (years) N = 14,193 N = 745 <0.01
CI 2.18.4) and that before 16 weeks 7.55 (95% CI 1.831.7) [6]. Mean (sd) 28.5 (W5.1) 29.04 (W5.1) 0.99
Assuming a causal relationship between bacterial vaginosis and <20 3.5% 2.3% <0.01
preterm delivery, with these odds ratios and a prevalence of 7% of 2024 22.5% 22.1%
2529 37.5% 32.5%
pregnant women exposed to bacterial vaginosis, the population
3034 25.1% 30.2%
attributable fraction due to BV would be 17% before 20 weeks and 3539 9.7% 11.1%
30% before 17 weeks. I40 years 1.7% 1.7%
In a recent meta-analysis [41], McDonald indicates that the Smoked during pregnancy N = 14,034 N = 731
Yes 23.4% 20.4% 0.06
treatment did not reduce the risk of PTB before 37 weeks or the
Educational level (n) N = 13,744 N = 719 0.9
risk of preterm prelabour rupture of membranes (PPROM). Completed Primary 4.8% 5.0%
However, treatment before 20 weeks gestation may reduce the Completed Secondary 35.8% 35.7%
risk of preterm birth less than 37 weeks. In women with a previous Post secondary 56.4% 55.8%
PTB, treatment did not affect the risk of subsequent PTB however it History of preterm delivery N = 13,165 N = 703
Yes 7.9% 9.7% 0.03
may decrease the risk of PPROM and low birth weight.
The U.S. Preventive Task Force (USPTF) does not recommend
screening for bacterial vaginosis in asymptomatic pregnant
References
women at low risk for preterm delivery. In asymptomatic pregnant
women at high risk for preterm delivery the current evidence is [1] Reid G. Probiotic Lactobacilli for urogenital health in women. Journal of Clinical
insufficient to assess the balance of benefits and harms of Gastroenterology 2008;42(Suppl. 3 Pt 2):S2346.

Please cite this article in press as: Desseauve D, et al. Prevalence and risk factors of bacterial vaginosis during the first trimester of
pregnancy in a large French population-based study. Eur J Obstet Gynecol (2012), http://dx.doi.org/10.1016/j.ejogrb.2012.04.007
G Model
EURO-7689; No. of Pages 5
D. Desseauve et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology xxx (2012) xxxxxx
[2] Donders GG, Van Calsteren K, Bellen G, et al. Predictive value for preterm birth
of abnormal vaginal flora, bacterial vaginosis and aerobic vaginitis during the
first trimester of pregnancy. BJOG 2009;116:131524.
[3] Hillier SL, Nugent RP, Eschenbach DA, et al. Association between bacterial
vaginosis and preterm delivery of a low-birth-weight infant. The Vaginal
Infections and Prematurity Study Group. New England Journal of Medicine
1995;333:173742.
[4] Banhidy F, Acs N, Puho EH, Czeizel AE. Rate of preterm births in pregnant
women with common lower genital tract infection: a population-based study
based on the clinical practice. Journal of Maternal-Fetal and Neonatal Medicine
2009;22:4108.
[5] Romero R, Gomez R, Chaiworapongsa T, et al. The role of infection in preterm
labour and delivery. Paediatric and Perinatal Epidemiology 2001;15(Suppl.
2):4156.
[6] Leitich H, Bodner-Adler B, Brunbauer M, Kaider A, Egarter C, Husslein P.
Bacterial vaginosis as a risk factor for preterm delivery: a meta-analysis.
American Journal of Obstetrics and Gynecology 2003;189:13947.
[7] McDonald HM, Brocklehurst P, Gordon A. Antibiotics for treating bacterial
vaginosis in pregnancy. Cochrane Database of Systematic Reviews
2007;CD000262.
[8] Gomez LM, Sammel MD, Appleby DH, et al. Evidence of a geneenvironment
interaction that predisposes to spontaneous preterm birth: a role for asymp-
tomatic bacterial vaginosis and DNA variants in genes that control the inflam-
matory response. American Journal of Obstetrics and Gynecology
2010;202(386):e16.
[9] Amsel R, Totten PA, Spiegel CA, Chen KC, Eschenbach D, Holmes KK. Nonspe-
cific vaginitis. Diagnostic criteria and microbial and epidemiologic associa-
tions. American Journal of Medicine 1983;74:1422.
[10] Meis PJ, Goldenberg RL, Mercer B, et al. The preterm prediction study:
significance of vaginal infections. National Institute of Child Health and
Human Development Maternal-Fetal Medicine Units Network. American
Journal of Obstetrics and Gynecology 1995;173:12315.
[11] Koumans EH, Sternberg M, Bruce C, et al. The prevalence of bacterial vaginosis
in the United States, 20012004: associations with symptoms, sexual beha-
viors, and reproductive health. Sexually Transmitted Diseases 2007;34:8649.
[12] Pastore LM, Thorp Jr JM, Royce RA, Savitz DA, Jackson TP. Risk score for
antenatal bacterial vaginosis: BV PIN points. Journal of Perinatology
2002;22:12532.
[13] Schwebke JR, Hillier SL, Sobel JD, McGregor JA, Sweet RL. Validity of the vaginal
gram stain for the diagnosis of bacterial vaginosis. Obstetrics and Gynecology
1996;88:5736.
[14] Trabert B, Misra DP. Risk factors for bacterial vaginosis during pregnancy
among African American women. American Journal of Obstetrics and Gyne-
cology 2007;197(477):e18.
[15] Cristiano L, Rampello S, Noris C, Valota V. Bacterial vaginosis: prevalence in an
Italian population of asymptomatic pregnant women and diagnostic aspects.
European Journal of Epidemiology 1996;12:38390.
[16] Goffinet F, Maillard F, Mihoubi N, et al. Bacterial vaginosis: prevalence and
predictive value of premature delivery and neonatal infection in women with
preterm labour and intact membranes. European Journal of Obstetrics Gyne-
cology and Reproductive Biology 2003;108:14651.
[17] Gratacos E, Figueras F, Barranco M, et al. Prevalence of bacterial vaginosis and
correlation of clinical to Gram stain diagnostic criteria in low risk pregnant
women. European Journal of Epidemiology 1999;15:9136.
[18] Jacobsson B, Pernevi P, Chidekel L, Platz-Christensen JJ. Bacterial vaginosis in
early pregnancy may predispose for preterm birth and postpartum endome-
tritis. Acta Obstetricia et Gynecologica Scandinavica 2002;81:100610.
[19] Kalinka J, Hanke W, Wasiela M, Laudanski T. Socioeconomic and environmen-
tal risk factors of bacterial vaginosis in early pregnancy. Journal of Perinatal
Medicine 2002;30:46775.
[20] Kurki T, Sivonen A, Renkonen OV, Savia E, Ylikorkala O. Bacterial vaginosis in
early pregnancy and pregnancy outcome. Obstetrics and Gynecology
1992;80:1737.
[21] Martnez de Tejada B, Coll O, de Flores M, Hillier SL, Landers DV. Prevalence of
bacterial vaginosis in an obstetric population of Barcelona. Medicina Clinica
1998;110:2014.
Please cite this article in press as: Desseauve D, et al. Prevalence and risk factors of bacterial vaginosis during the first trimester of
pregnancy in a large French population-based study. Eur J Obstet Gynecol (2012), http://dx.doi.org/10.1016/j.ejogrb.2012.04.007
View publication stats
5
[22] Nicand E, Cavallo JD, Crenn Y, Meyran M. Value of the score for Gram strains in
the diagnosis of bacterial vaginosis. Pathologie Biologie 1994;42:53943.
[23] Svare JA, Schmidt H, Hansen BB, Lose G. Bacterial vaginosis in a cohort of
Danish pregnant women: prevalence and relationship with preterm delivery,
low birthweight and perinatal infections. BJOG 2006;113:141925.
[24] Thorsen P, Vogel I, Molsted K, et al. Risk factors for bacterial vaginosis in
pregnancy: a population-based study on Danish women. Acta Obstetricia et
Gynecologica Scandinavica 2006;85:90611.
[25] Ugwumadu A, Manyonda I, Reid F, Hay P. Effect of early oral clindamycin on
late miscarriage and preterm delivery in asymptomatic women with abnormal
vaginal flora and bacterial vaginosis: a randomised controlled trial. Lancet
2003;361:9838.
[26] Hay PE, Morgan DJ, Ison CA, et al. A longitudinal study of bacterial vaginosis
during pregnancy. British Journal of Obstetrics and Gynaecology
1994;101:104853.
[27] Lamont RF, Morgan DJ, Wilden SD, Taylor-Robinson D. Prevalence of bacterial
vaginosis in women attending one of three general practices for routine
cervical cytology. International Journal of STD and AIDS 2000;11:4958.
[28] Ralph SG, Rutherford AJ, Wilson JD. Influence of bacterial vaginosis on con-
ception and miscarriage in the first trimester: cohort study. BMJ
1999;319:2203.
[29] Peipert JF, Lapane KL, Allsworth JE, Redding CA, Blume JD, Stein MD. Bacterial
vaginosis, race, and sexually transmitted infections: does race modify the
association? Sexually Transmitted Diseases 2008;35:3637.
[30] Larsson PG, Fahraeus L, Carlsson B, Jakobsson T, Forsum U. Predisposing factors
for bacterial vaginosis, treatment efficacy and pregnancy outcome among
term deliveries; results from a preterm delivery study. BMC Womens Health
2007;7:20.
[31] Nugent RP, Krohn MA, Hillier SL. Reliability of diagnosing bacterial vaginosis is
improved by a standardized method of gram stain interpretation. Journal of
Clinical Microbiology 1991;29:297301.
[32] Bresson L, Massoni S, Jailloux-Beaurain C, et al. Self-collected vaginal swabs to
diagnosis bacterial vaginosis during pregnancy: a pilot study. Gynecologie
Obstetrique et Fertilite 2006;34:7015.
[33] University Hospital, Lille. Prevention of Very Preterm Delivery by Testing for
and Treatment of Bacterial Vaginosis (PREMEVA); 2008, http://clinicaltrials.-
gov, NCT00642980.
[34] Chaix B, Chauvin P. The contribution of multilevel models in contextual
analysis in the field of social epidemiology: a review of literature. Revue
dEpidemiologie et de Sante Publique 2002;50:48999.
[35] Goldenberg RL, Klebanoff MA, Nugent R, Krohn MA, Hillier S, Andrews WW.
Bacterial colonization of the vagina during pregnancy in four ethnic groups.
Vaginal Infections and Prematurity Study Group. American Journal of Obstet-
rics and Gynecology 1996;174:161821.
[36] Nelson DB, Bellamy S, Gray TS, Nachamkin I. Self-collected versus provider-
collected vaginal swabs for the diagnosis of bacterial vaginosis: an assessment
of validity and reliability. Journal of Clinical Epidemiology 2003;56:8626.
[37] Goffinet F. Antenatal and peripartum antibiotic therapy in the case of prema-
ture rupture of the membranes. Journal de Gynecologie Obstetrique et Biologie
de la Reproduction 1999;28:6509.
[38] Pavlova SI, Tao L. Induction of vaginal Lactobacillus phages by the cigarette
smoke chemical benzo[a]pyrene diol epoxide. Mutation Research 2000;466:
5762.
[39] Lejeune VN, Chaplet VM, Carbonne B, Jannet DJ, Milliez JM. Precarity and
pregnancy in Paris. European Journal of Obstetrics Gynecology and Reproduc-
tive Biology 1999;83:2730.
[40] Leitich H, Kiss H. Asymptomatic bacterial vaginosis and intermediate flora as
risk factors for adverse pregnancy outcome. Best Practice and Research Clinical
Obstetrics and Gynaecology 2007;21:37590.
[41] McDonald HM, Brocklehurst P, Gordon A. Antibiotics for treating bacterial
vaginosis in pregnancy. Cochrane Database of Systematic Reviews 2007;(1):
CD000262.
[42] U.S. Preventive Services Task Force. Screening for bacterial vaginosis in
pregnancy to prevent preterm delivery: U.S. Preventive Services Task Force
recommendation statement. Annals of Internal Medicine 2008;148:2149.