1. Arch Gynecol Obstet (2010) 281:589–600
DOI 10.1007/s00404-009-1318-3
MATERNO-FETAL MEDICINE
Vaginal microbial flora and outcome of pregnancy
Laura Donati • Augusto Di Vico • Marta Nucci • Lorena Quagliozzi •
Terryann Spagnuolo • Antonietta Labianca • Marina Bracaglia •
Francesca Ianniello • Alessandro Caruso • Giancarlo Paradisi
Received: 13 May 2009 / Accepted: 26 October 2009 / Published online: 5 December 2009
Ó Springer-Verlag 2009
Abstract Conclusions Vaginal ecosystem study with the detection
Background The vaginal microflora of a healthy asymp- of pathogens is a key instrument in the prevention of pre-
tomatic woman consists of a wide variety of anaerobic and term delivery, pPROM, chorioamnionitis, neonatal,
aerobic bacterial genera and species dominated by the puerperal and maternal-fetal infections.
facultative, microaerophilic, anaerobic genus Lactobacil-
lus. The activity of Lactobacillus is essential to protect Keywords Vaginal flora Á Bacterial vaginosis Á
women from genital infections and to maintain the natural Clinical diagnosis Á Pregnancy
healthy balance of the vaginal flora. Increasing evidence
associates abnormalities in vaginal flora during pregnancy
with preterm labor and delivery with potential neonatal Introduction
sequelae due to prematurity and poor perinatal outcome.
Although this phenomenon is relatively common, even in The vaginal microflora of healthy asymptomatic women
populations of women at low risk for adverse events, the consists of a wide variety of anaerobic and aerobic bacterial
pathogenetic mechanism that leads to complications in genera and species dominated by the facultative, microaer-
pregnancy is still poorly understood. ophilic, anaerobic genus Lactobacillus [1, 2]. Lactobacilli
Objective This review summarizes the current knowledge are the most well-known markers of normal vaginal flora.
and uncertainties in defining alterations of vaginal flora in The activity of Lactobacillus is essential to protect
non-pregnant adult women and during pregnancy, and, in women from genital infections and to maintain the natural
particular, investigates the issue of bacterial vaginosis and healthy balance of the vaginal flora. This role is particu-
aerobic vaginitis. This could help specialists to identify larly important during pregnancy because vaginal infection
women amenable to treatment during pregnancy leading to has been claimed as one of the most important mechanisms
the possibility to reduce the preterm birth rate, preterm responsible for preterm birth and perinatal complications.
premature rupture of membranes, chorioamnionitis, neo- In normal conditions, Lactobacillus spp. utilizing available
natal, puerperal and maternal–fetal infectious diseases. glycogen produce lactic acid, which is able to acidify the
vaginal pH to less than 4.5, inhibiting the growth of non-
acid tolerant microorganisms, known as potentially path-
ogenic. Moreover, Lactobacillus spp. also produce hydro-
L. Donati Á A. Di Vico Á M. Nucci Á L. Quagliozzi Á gen peroxide, a potent antimicrobial molecule, toxic to
T. Spagnuolo Á A. Labianca Á M. Bracaglia Á F. Ianniello Á
other microorganisms [3]. There are many different strains
A. Caruso Á G. Paradisi
Department of Obstetrics and Gynecology, Catholic University of lactobacilli present in the vagina, the most frequent
of Sacred Heart, Largo A. Gemelli 8, 00168 Rome, Italy being L. jensenii, L. gasseri, L. iners and L. crispatus, and
there is a wide variation in species and relative numbers of
G. Paradisi (&)
species according to the population studied. When ‘‘nor-
Department of Obstetrics and Gynecology, Catholic University
of Sacred Heart, Via Servilio IV 4, 00178 Rome, Italy mal’’, Bacterial flora is predominantly lactobacillary type.
e-mail: giancarlo.paradisi@tin.it When ‘‘abnormal’’, the flora can be disturbed by anaerobic
123

2. 590 Arch Gynecol Obstet (2010) 281:589–600
overgrowth [bacterial vaginosis (BV)] or by aerobic that of White women among subjects who were not diag-
microorganisms such as Escherichia coli, group B strep- nosed as having BV [10]. Additional studies suggested that
tococci, enterococci, etc. [aerobic vaginitis (AV)], or can this difference was only statistically significant among
be a mixture of both (mixed abnormal flora). women who had abnormal, i.e., non-lactobacilli-domi-
Therefore, it is clearly evident that where lactobacilli nated, vaginal microflora [11, 12].
predominate, other bacteria and parasites such as Tricho- Regardless of the predominant bacterial species in the
monas are not abundant. On the other hand, Lactobacillus- vagina of a healthy premenopausal woman, it appears
deficient conditions are associated with the development of certain that lactic acid production is crucial to the main-
numerous infectious conditions such as the previously tenance of a healthy vaginal ecosystem. The resulting
mentioned BV and AV, and promote the transmission of acidic pH prevents the overgrowth of potentially patho-
sexually transmitted diseases such as Gonorrhoea, Chla- genic microorganisms. Additional benefits for the host
mydia, Syphilis, Trichomoniasis, HIV and HPV. Normal of Lactobacillus predominance are the production of
and abnormal lactobacillary flora is divided into three or hydrogen peroxide and bacteriocins by strains of these
four floral types also called as ‘‘Lactobacillary grades’’. microbes.
Normal—grade I—flora corresponds to predominantly The composition of the vaginal ecosystem is not static
lactobacillary morphotypes, with very few coccoid bacteria but changes over time and in response to endogenous and
present. The intermediate—grade II—flora corresponds to exogenous influences [13]. Variables include the stage of
a diminished lactobacillary flora, mixed with other bacte- the menstrual cycle, pregnancy, use of contraceptive
ria. This group is subdivided into slightly disturbed, fairly agents, frequency of sexual intercourse, specific sexual
normal (IIa) and moderately disturbed, rather abnormal partners, vaginal douching, use of panty liners or vaginal
(IIb) lactobacillary flora. Finally, the grossly abnormal— deodorants, and the utilization of antibiotics or other
grade III—flora consists of numerous other bacteria, with medications with immune or endocrine activities.
no lactobacilli present. In order to diagnose such abnormal During the menstrual cycle, vaginal levels of hormones
lactobacillary grades, the use of the wet mount is preferred and glycogen vary and menstrual blood alters vaginal pH
to the Gram stain due to its superior accuracy and better and provides a substrate for many microorganisms. Nev-
correlation with vaginal lactate, accepted by most as the ertheless, levels of vaginal lactobacilli appear to remain
best functional test for lactobacillary defense function. constant throughout the cycle; non-Lactobacillus species
increase during the proliferative phase, while Candida
albicans concentrations are highest towards menstruation
Normal vaginal flora (as determined by culture) [14].
As previously stated, the dominant species of vaginal flora
is initially identified as Lactobacillus Acidophilus. Bacterial vaginosis
In healthy women with a Lactobacillus-dominant vagi-
nal microflora, the major phylotypes detected by gene Bacterial vaginosis, as previously mentioned, is charac-
amplification are L. crispatus and L. iners [4–7] or L. terized by a change in vaginal microflora in which the
crispatus and L. gasseri [6]. Additional species, such as L. normally occurring lactobacilli yield to an overgrowth of a
jensenii, L. gallinarum and L. vaginalis, have also been mixed anaerobic bacterial flora (Fig. 1). The most common
identified in some women. Interestingly, apparently healthy symptom among women with BV is a thin, gray, non-
vaginal ecosystems are maintained in some women in the pruritic discharge with a fishy odor, while the vagina is not
absence of a Lactobacillus-dominant vaginal microflora. red or inflamed and there is no prominent symptoms of
Atopobium, Megasphaera and Leptotrichia are all pro- burning, pain or dyspareunia. Thus, BV is usually a
ducers of lactic acid [4, 8] similar to the lactobacilli. polymicrobial infection, and organisms responsible for
Therefore, the acidic environment of the vagina, recog- infection include: Gardnerella vaginalis, Mycoplasma
nized as an important defense mechanism against the hominis, Bacteroides spp., Peptostreptococcus spp., Fuso-
proliferation of different microbial pathogens, can be bacterium spp., Prevotella [15, 16].
maintained by bacterial species other than the lactobacilli. A typical feature of BV is the absence of inflammation. In
Possible racial/ethnic differences in the composition of BV, there is only a slight increase in interleukin I and an
the ‘‘normal’’ microflora have also not yet received ade- unexpectedly low production of interleukin 8, preventing
quate research attention. The occurrence of hydrogen-per- the attraction of inflammatory cells such as macrophages
oxide-producing lactobacilli, active in antimicrobial and neutrophils. Hence, if severe inflammation is present
defense, is lower in Black women [9]. It has also been (e.g., when more than ten leukocytes are present per epi-
noted that the vaginal pH of Black women is higher than thelial cell), another diagnosis has to be considered. Indeed,
123

3. Arch Gynecol Obstet (2010) 281:589–600 591
G. vaginalis on BV. However, since G. vaginalis was also
detected in a minority of asymptomatic women, it is clear
that its presence ‘‘per se’’ is not sufficient for the devel-
opment of BV [25].
Mycoplasmas and ureaplasmas are eubacteria. They are
the smallest self-replicating organisms, both in genome size
and cellular dimensions. Mycoplasma and Ureaplasma
species are included within the class Mollicutes. Three
species in the genus Mycoplasma, M. hominis, M. genita-
lium and M. fermentans, are known to occur in the female
urogenital tract and have been associated with disease.
However, ureaplasmas are unique among the Mollicutes in
that they hydrolyze urea to generate metabolic energy. The
cytadherence proteins of M. hominis and Ureaplasma spp.
Fig. 1 Bacterial vaginosis are not organized into a demonstrable attachment organelle
and they have not been completely characterized. The
concomitant cervicitis, trichomoniasis, candidiasis and AV M. hominis variable adherence-associated (Vaa) antigen is a
are all known to present with an increased immune response, size- and phase-variable adhesin which is highly immuno-
with increased numbers of monocytes and leukocytes in the genic. The high variability of Vaa is presumably important
vast majority of cases. Recently, due to the heterogeneity of for the diversity and host adaptation of this mycoplasma.
microorganisms involved in the pathogenesis and the clin- Similar to the Vaa of M. hominis, the multiple-banded (MB)
ical findings, several efforts have been made to define antigen of Ureaplasma spp. is immunogenic, undergoes a
unequivocally the presence of an abnormal vaginal flora and high rate of variation in vitro, may be involved in the
to standardize the diagnosis and therapy. The Amsel criteria stimulation of the host inflammatory response, and is vari-
and scoring of Gram-stained smears of vaginal fluid are the able in size on invasive isolates. Ureaplasmas attach to host
most commonly used diagnostic principles for BV. The four erythrocytes, neutrophils, spermatozoa and urethral epi-
Amsel criteria are usually regarded as the gold standard. thelial cells, and they can directly activate the first com-
Criteria include: the presence of a thin, homogeneous dis- ponent of complement. Both M. hominis and Ureaplasma
charge, which adheres to the vaginal walls, vaginal pH spp. can induce inflammation in humans and this is a major
above 4.5, release of fishy odor upon alkalinization with factor involved in the production and manifestation of
10% potassium hydroxide and clue cells on a saline well clinical disease. Secretory products such as ammonia gen-
mount. If three of these four criteria are met, the patient is erated from the metabolism of arginine by M. hominis and
diagnosed with BV [17]. However, these criteria are often urea by Ureaplasma spp. may produce a local cytotoxic
modified, as the typical homogenous discharge (the most effect. Urease production by ureaplasmas has been impli-
subjective sign) is often not included. Among the organisms cated in urinary calculus production [26].
responsible for infection, G. vaginalis and Mycoplasma are
the most important and the most common.
In the 1950s, Leopold [18] and then Gardner and Dukes The ‘‘intermediate flora’’
[19] observed abundant small, pleomorphic gram-variable
rods in the genital tracts of women with BV. This organ- Ideally, the ‘‘ intermediate flora’’ state represents a turning
ism, first called Haemophilus vaginalis [20] and then point from a normal state into BV, or from BV to normal
repeatedly renamed, is now classified as G. vaginalis, the (Fig. 2). In reality, however, most of the women with so-
sole member of the genus Gardnerella [21, 22]. Phyloge- called intermediated BV according to Nungent will neither
netic analysis based on 16S rRNA places Gardnerella in have BV nor a normal flora. It is becoming clear that the
the Gram-positive family Bifidobacteriales. An abundance intermediate group is linked to a different and usually more
of G. vaginalis and a paucity of Lactobacillus species are serious range of complications, including mid-trimester
characteristic of a BV-associated microflora, but the rela- pregnancy loss, than the ‘‘classic’’ full-blown BV [27–29].
tive contribution of G. vaginalis to the pathogenesis of BV The criteria for the diagnosis of intermediate vaginal flora is
is not yet clear. Recently, in a series of clinical studies, Gram-based; there is also the definition of Spiegel et al. [30]
Swidsinski et al. demonstrated that G. vaginalis has the in which the intermediate category is defined as the pres-
unique characteristic to develop an adherent biofilm, spe- ence of reduced lactobacilli mixed with other morphotypes
cific for BV, that persists after metronidazole treatment [27] or it is defined as the presence of intermediate
[23, 24]. These findings added new insight into the role of microbial flora, predominantly M. hominis, Ureaplasma
123

4. 592 Arch Gynecol Obstet (2010) 281:589–600
– Presence of parabasal epithelial cells.
– The type of background flora.
Parabasal cells were added because they were con-
sidered to be a sign of inflammation usually not seen in
uncomplicated BV, but rather in severe forms of AV,
such as in desquamative inflammatory vaginitis [33].
Background flora was allocated score 0 if the back-
ground flora was unremarkable or showed debris and
bare nuclei from lysed epithelial cells (cytolysis) [34],
score 1 if the lactobacillary morphotypes were very
coarse or resembled small bacilli (other than lactobacilli),
and 2 if there were prominent cocci, or chained cocci
visible. A composite score of 1–2 represents normality.
Fig. 2 Intermediate vaginosis
A score of 3–4 corresponds to slight AV, a score of 5–6
to moderate vaginitis, and a score above 6 (to maximum
urealyticum, G. vaginalis, Gram-negative anaerobic rods,
10) to severe AV. In practice, a score of 8–10 is usually
Chlamydia trachomatis and few Lactobacillus spp. [31].
identical to so-called ‘‘desquamative vaginitis’’, so that
Leitich H. and Kiss H., in their meta-analysis, evaluated
such a diagnosis can be seen as the most extreme form
BV and intermediate vaginal flora as risk factors for
of AV [33, 35].
adverse pregnancy outcome (Table 1). Fourteen new
The question remains if AV is an entity that really dif-
studies with results for 10,286 patients are included, so the
fers from BV. The pro-inflammatory cytokines produced in
results for 30,518 patients in 32 studies are available for
the host’s vagina should also differ between the two enti-
this meta-analysis. Intermediate vaginal flora is not sig-
ties, as full-blown BV is typically devoid of any host leu-
nificantly associated with any outcome included.
kocytosis response, while AV is a real vaginitis. Indeed,
major differences in response were seen between AV and
BV. Interleukin-1-b increases in both conditions, but sig-
Aerobic vaginitis
nificantly more so in AV (eightfold): vaginal IL-1-b con-
centrations increase as lactobacilli decrease [36]. IL-1-b
In recent years, Donders et al. [32] concentrated their
tends to increase even more dramatically when the
attention on women with abnormal vaginal flora with or
inflammatory response of the host increases. In a study by
without concomitant BV and described a clinical and
Mattsby-Baltzer et al. [37] in women during their first
microscopic pattern that fits a new condition defined as
trimester of pregnancy, BV was associated with IL-1-b, but
‘‘AV’’, raising the possibility that there may be a degree of
not with IL-6. Similarly, in another study of women in
overlap between these two entities, indicating that AV and
labor, IL-1b was associated with BV, but IL-6 was not [38].
BV can coexist.
IL-6 remained unchanged in women with BV, when
We are convinced that AV should be seen as a separate
compared to women with normal flora. In women with AV,
disease entity.
on the other hand, IL-6 is increased fivefold. IL-6 is a well-
Clinical features associated with them and the host
known marker for bacterial amnionitis and imminent term
response to AV are so specific for the conditions and differ
and preterm delivery [38, 39]. IL-6 and IL-8 are known
clearly from BV. Further studies to differentiate the effects
chemo-attractants and are directly linked to increased
of BV and AV on the outcome of pregnancy are therefore
prostaglandin production and delivery [38]. Finally, the
urgently needed, as they may hold part of the answer to the
production of LIF, a relatively unknown cytokine, is
question why some studies have found no association
threefold lower in women with BV, but threefold higher in
between BV or its treatment and pregnancy outcome, while
women with AV. Even though these differences are not
others have found that restoring the flora to normal pre-
quite statistically significant, vaginal LIF concentration
vented preterm birth.
may be seen as a marker for AV.
Diagnosis of AV is based on microscopy (Fig. 3).
There is an evident correlation between AV and group B
Lactobacillary grades (see above) are the basis for a
streptococci, S. aureus and E. coli. The more severe the
composite score to which any of the four following vari-
microscopic findings, the more likely these organisms will
ables is added:
be cultured. Case studies have recently drawn attention to
– Proportional number of leukocytes. the possibility of vaginitis due to group B streptococci [40],
– Presence of toxic leukocytes. and Monif provides in vitro evidence that group B
123

5. Table 1 Intermediate vaginal flora and adverse pregnancy outcome: patients, method, results of individual studies
Study Patient inclusion Diagnosis of Mean gestational Outcome Patients with Patients with OR (95%CI)
criteria intermediate age at BV screening intermediate normal vaginal
vaginal flora (weeks) vaginal flora flora
Donders et al. [66] Singleton pregnancy at Clinical criteria 9.1 Delivery 37 weeks 5/10 20/180 8.00 (2.13–30.06)
14 weeks (B Amsel criteria) and Late miscarriage 20 weeks 1/11 12/192 1.50 (0.18–12.71)
abnormal vaginal
Arch Gynecol Obstet (2010) 281:589–600
bacterial flora
Edwards et al. [67] Singleton pregnancy at Gram stain (Nungent 28.3 Delivery 37 weeks 5/19 20/86 1.18 (0.38–3.67)
23–32 weeks and score 4–6)
preterm labor
Goffinet et al. [68] Pregnancy at Gram stain (Nungent 29.0 Delivery 35 weeks 21/76 50/254 1.56 (0.86–2.81)
24–34 weeks and score 4–6) Delivery 33 weeks 14/76 27/254 1.90 (0.94–3.84)
preterm labor
Chorioamnionitis 3/76 4/254 2.57 (0.56–11.74)
Neonatal infection 6/76 13/254 1.59 (0.58–4.33)
Guerra et al. [69] Singleton pregnancy at Gram stain (Nungent 8.5 Delivery 37 weeks 24/48 17/90 6.01 (2.76–13.11)
10 weeks score 4–6) Late miscarriage B25 weeks 4/52 5/95 1.50 (0.38–5.85)
25.0 Delivery 37 weeks 17/23 43/145 6.72 (2.48–18.21)
Hay et al. [27] Singleton pregnancy at Gram stain (Spiegel 12.5 Delivery 37 weeks 0/19 9/380 0.0 (n.c.)
9–24 weeks definition with Late miscarriage 1/20 4/384 5.00 (0.53–46.93)
addition of an
At 16–24 weeks
intermediate category)
16.5 Delivery 37 weeks 1/32 17/584 1.08 (0.14–8.35)
Kalinka et al. [31] Singleton pregnancy at Gram stain (Spiegel 12.3 Delivery 37 weeks 8/71 6/70 1.35 (0.44–4.13)
8–16 weeks definition with
addition of an
intermediate category)
Oakeshott et al. [70] Singleton pregnancy at Gram stain (Nungent 7.0 Delivery 37 weeks 1/39 46/746 0.40 (0.05–2.98)
10 weeks score 4–6) Late miscarriage 2/41 8/754 4.78 (0.98–23.28)
At 13–23 weeks
Modified from Leitich H. and Kiss H. (2007)
593
123

6. 594 Arch Gynecol Obstet (2010) 281:589–600
Neisseria gonorrhoeae, C. trachomatis and Trichomonas
vaginalis.
Haemophilus influenzae only occasionally acts like a
pathogen, being associated with preterm labor. Group B
hemolytic Streptococcus can have devastating effects on
the preterm or low birth weight infant.
Bacterial vaginosis is a polymicrobial condition
increasingly associated with adverse perinatal sequelae. No
single organism causes BV; however, there appears to be
an independent association between BV and G. vaginalis,
Mobiluncus spp., anaerobic Gram-negative rods, and M.
hominis. The exact role and significance that M. hominis
and Ureaplasma spp. play in BV remain uncertain [25].
The reason why BV causes preterm birth or labor in some
Fig. 3 Aerobic vaginosis women, and remains practically paucisymptomatic and
without any complications for pregnancy in others, is still
streptococci inhibit the growth of lactobacilli and G. vag- poorly understood.
inalis, but not S. aureus [41, 42].
Risk in pregnancy
Vaginal flora and pregnancy
Vaginal infection represents one of the most important risk
During physiological pregnancy, the higher levels of factors for complications of pregnancy such as premature
estrogens induce not only a better epithelial tropism but rupture of membrane, preterm labor and birth and perinatal
also a positive effect on lactobacillary activity and prolif- infection. In the following paragraph, we have tried to
eration due to an increased glycogen availability. summarize recently obtained information in this field.
As pregnancy advances, the genital tract flora becomes
progressively more benign, until term. It is particularly Preterm delivery and late miscarriage
difficult to define abnormal genital tract flora in pregnancy.
Usui et al. [43] prospectively examined the vaginal flora Bacterial vaginosis occurs in up to 20% of pregnant women
in a sample of 1,958 pregnant women during the first and and has been associated with premature birth and sponta-
second trimesters and analyzed the relative percentage of neous abortion [44]. This condition is characterized by a
each bacterial species found. Aerobic Gram-negative rods watery discharge with a fishy odor, but half of the women
and cocci, anaerobic Gram-negative rods, and lactobacilli with this infection may be asymptomatic or experience
were analyzed using standard laboratory methods. In only mild symptoms [45]. The natural history of BV is
addition, the percentage of women with preterm birth was such that it may spontaneously resolve without treatment
analyzed: 6% of women delivered before the 33rd week although most women identified as having BV in early
(n = 118) and 11% before the 37th week (n = 224). In pregnancy are likely to not have persistent infection later in
these cohort of patients, Lactobacilli were significantly pregnancy [46]. Increasing evidences associates BV in
decreased, whereas the presence of other bacteria was pregnancy with poor perinatal outcome, in particular an
greater in patients with the absence of Lactobacilli. The increased risk of preterm birth with potential neonatal
Mycoplasmas were not influenced by the presence of sequelae due to prematurity.
Lactobacilli flora. The mechanism by which BV may lead to preterm
These results indicate that the element that exposes the labor/preterm birth has not yet been well defined. A pos-
risk of preterm birth seems to be the absence of Lactoba- sible explanation involves alterations in the host defense
cilli, rather than the presence of other microorganisms and mechanism that leads to ascending intrauterine infection.
suggests that a test for determining the presence of vaginal Since individuals differ in their innate ability to mount an
lactobacilli may be a clinically useful tool for identifying inflammatory response to bacterial products, it has been
women at risk for preterm delivery at 33 weeks of proposed that women who are immunologically hypore-
gestation. sponsive may not be able to control a large bacterial attack
The numerous types of organisms which can be found in and may be predisposed to ascending infection with vari-
association with preterm labor and delivery can be classi- able fetal consequences [47]. On the other hand, women
fied into four main groups. There are three main pathogens: who respond to a bacterial stimulus with exaggerated
123

7. Arch Gynecol Obstet (2010) 281:589–600 595
cytokine production at the maternal–fetal interface may be Bacterial morphotypes and pregnancy complications
at greater risk of preterm labor if microorganisms gain
access to the choriodecidual space. Hillier et al. [48] Table 3 presents the data regarding the specific bacterial
reported that BV was significantly associated with the morphotypes present by trimester. Consistent with
isolation of microorganisms from the chorioamnion, but Waters’s findings of Nugent score by trimester, there was
was unable to determine the effects of individual bacterial an observed trend toward a Gram stain without any
species, or to address the question of whether BV is abnormalities as pregnancy progressed [54]. Some micro-
associated with premature delivery independently of organisms, due to their higher incidence in the population,
chorioamnion infection. deserve clarification: while a number of genital microor-
BV more than doubled the risk of preterm delivery in ganisms such as E. coli, Listeria monocytogenes and viri-
asymptomatic patients (OR 2.31, 95% CI: 1.56–3.00) and dans streptococci may be involved in chorioamnionitis,
in patients with symptoms of preterm labor (OR 2.38, 95% harboring of these organisms during early- to mid-preg-
CI 1.02–5.58). BV also significantly increased the risk of nancy has not been associated with an increased risk of
late miscarriages (OR 6.32, 95% CI 3.65–10.94) and preterm labor [55, 56].
maternal infection (OR 2.53, 95% CI 1.26–5.08) in Mycoplasma hominis and Ureaplasma species have
asymptomatic patients (Table 2). been associated with a variety of conditions that may affect
These results confirm that BV is a risk factor for preterm the gravida, the developing fetus, and the neonate. For the
delivery and maternal infectious morbidity and a strong gravida and fetus, these organisms may contribute to pre-
risk factor for late miscarriage [49]. mature labor, chorioamnionitis, postpartum endometritis,
Nearly one quarter of pregnant white women in a growth restriction, spontaneous abortion and stillbirth;
National Health and Nutrition Examination Survey while the exposed neonate may develop pneumonia, bac-
(NHANES) probability vaginal sample had Gram stains teremia, meningitis, abscesses and chronic lung disease.
consistent with BV [50]. BV in pregnancy may be more Intrauterine infections may trigger premature labor and
common among minority women, those of low socioeco- lead to preterm birth. The mechanisms by which intra-
nomic status, and those who have previously delivered low uterine infections lead to preterm labor are related, as
birth weight infants. The National Institute of Child Health mentioned before, to the activation of the innate immune
and Human Development Maternal–Fetal Medicine Units system. Microorganisms are recognized by pattern-recog-
Network study found that nearly 50% of pregnant African- nition receptors (e.g., Toll-like receptors), which in turn
American women had BV, similar to the rate found in non- elicit the release of inflammatory chemokines and cyto-
pregnant African-American women in NHANES [51]. BV kines. These cytokines, elaborated at the maternal–fetal
is relatively common, even in populations of women at low interface, trigger prostaglandin production in the amnion,
risk of adverse events, and as it is amenable to appropriate chorion, decidua and myometrium, leading to uterine
therapy [52, 53], identification during pregnancy and contractions, cervical dilatation, membrane rupture and
treatment may theoretically represent a rare opportunity to uterine contractions which further facilitate bacterial entry
reduce the preterm birth rate, resulting risk of prematurity into the uterine cavity. Intra-amniotic infection contributes
to the newborn. Such treatment may also reduce other to 40% of peripartum febrile illness and is associated with
adverse perinatal outcomes such as postpartum infection. at least one-third of early-onset neonatal sepsis. The inci-
Moreover, BV is diagnosed mostly in the first trimester dence increases with decreasing gestational age at delivery
and the prevalence decreases in the second and third. [57]. The prevalence of positive cultures and bacterial
Interestingly, only 9.4% of gravid patients have a positive DNA in choriodecidual tissues can be greater than in
diagnosis for BV, according to the Nugent criteria, in all amniotic fluid, lending further support to the idea that
three trimesters, regardless of the treatment [32]. These microorganisms ascend from the vagina through the chor-
findings suggest that some BV are self-limited and com- ioamniotic space to gain access to the amniotic cavity and,
plication-free. However, it is also known that 20% do not subsequently, the fetus. There is also evidence that bacte-
progress to a normal Nugent score by the third trimester ria, including ureaplasmas, may colonize the endometrial
and approximately 15% were persistently BV intermediate cavity prior to implantation [58]. Ureaplasma spp. are the
(Nugent 4–6). microorganisms most frequently isolated from amniotic
Interestingly, in the setting of BV during pregnancy, the fluid or placentae in women who deliver prematurely,
incidence of different bacterial morphotypes varies either with preterm premature rupture of membranes or in
between the three trimesters (Table 2), demonstrating a preterm labor with intact membranes [59]. These organ-
dynamic phenomenon and/or different types of imbalance isms have been isolated in the amniotic fluid as early as the
in vaginal bacterial flora during different phases of 16th week and can result in a clinically silent chronic and
gestation. progressive infection where delivery does not occur for
123

8. 596 Arch Gynecol Obstet (2010) 281:589–600
Table 2 Bacterial vaginosis and adverse pregnancy outcome: patients, methods and results of individual studies
Study Patient inclusion Diagnosis of BV Mean Outcome Patients Patients OR (95%CI)
Criteria gestational with BV without
age at BV BV
screening
(weeks)
Andrews et al. [71] Singleton pregnancy at Gram stain 23.7 Delivery 37 weeks 15/99 23/217 1.51 (0.75–3.03)
(RCT) 21–25 weeks and (Nungent
positive fetal score C7)
fibronectin test
Daskalakis et al. [72] Singleton pregnancy at Gram stain 23.5 Delivery 37 weeks 16/95 88/1102 2.33 (1.31–4.17)
22–25 weeks and no (Nungent
previous preterm score C7) and
delivery vaginal pH
[4.5
De Seta et al. [73] Singleton pregnancy at Gram stain 15.5 Delivery 37 weeks 14/95 35/503 2.31 (1.19–4.49)
13–18 weeks (Nungent
score C7)
Edwards et al. [67] Singleton pregnancy at Gram stain 28.3 Delivery 37 weeks 9/23 25/105 2.06 (0.80–5.32)
23–32 weeks and (Nungent
preterm labor score C7)
Genc et al. [74] Singleton pregnancy at Gram stain 20.0 Delivery 37 weeks 4/30 19/177 1.28 (0.40–4.06)
18–22 weeks and no (Nungent
previous preterm score C7)
delivery 24 weeks
Goffinet et al. [68] Pregnancy at Gram stain 29.0 Delivery 35 weeks 6/24 71/330 1.22 (0.47–3.18)
24–34 weeks and (Nungent Delivery 33 weeks 6/24 41/330 2.35 (0.88–6.26)
preterm labor score C 7)
Chorioamnionitis 1/24 7/330 2.01 (0.24–17.01)
Neonatal infection 2/24 19/330 1.49 (0.33–6.80)
Goyal et al. [75] Pregnancy at Gram stain 30.2 Delivery 37 weeks 18/19 30/41 6.60 (0.79–55.48)
23–36 weeks and (Nungent
preterm labor score C7)
Guerra et al. [69] Singleton pregnancy at Gram stain 8.5 Delivery 37 weeks 33/72 45/138 1.75 (0.97–3.14)
10 weeks and C1 (Nungent 25.0 Late miscarriage 23/95 9/147 4.90 (2.15–11.14)
previous preterm score C7) B25 weeks
delivery
Delivery 37 weeks 12/36 60/168 0.90 (0.42–1.93)
Kalinka et al. [31] Singleton pregnancy at Gram stain 12.3 Delivery 37 weeks 9/55 14/141 1.77 (0.72–4.38)
8–16 weeks (Nungent
score C7)
Kalinka et al. [76] Singleton pregnancy at Gram stain 29.0 Delivery 37 weeks 5/31 10/83 1.40 (0.44–4.49)
22–34 weeks (Nungent
score C7)
Kiss et al. [77] (RCT) Singleton pregnancy at Gram stain 17.0 Delivery 37 weeks 10/179 116/1918 0.92 (0.47–1.79)
15–19 weeks (Nungent
score C7)
Oakeshott et al. [70] Singleton pregnancy at Gram stain 7.0 Delivery 37 weeks 7/112 47/785 1.05 (0.46–2.38)
10 weeks (Nungent Late miscarriage at 5/117 10/795 3.50 (1.18–10.44)
score C7) 13–23 weeks
Purwar et al. [78] Singleton pregnancy at Gram stain 25.5 Delivery 37 weeks 32/115 40/823 7.55 (4.50–12.66)
12–28 weeks (Nungent
score C7)
Thorsen et al. [79] Pregnancy at Gram stain 17.0 Delivery 37 weeks 13/401 99/2526 0.82 (0.46–1.48)
24 weeks (Nungent Delivery 34 weeks 7/401 39/2526 1.13 (0.50–2.55)
score C7)
Delivery 32 weeks 2/401 22/2526 0.57 (0.13–2.44)
Modified from Leitich H. and Kiss H. (2007)
several weeks [60]. In a recent study of 254 asymptomatic correlated with subsequent preterm labor and preterm
women, the detection of the Ureaplasma species by PCR delivery [59]. It is correct to emphasize that physiological
assay in second trimester amniotic fluid was highly colonization of the vagina with lactobacilli appears to play
123

9. Arch Gynecol Obstet (2010) 281:589–600 597
Table 3 Bacterial morphotypes by trimester
Gram stain I trimester II trimester III trimester
(n = 148), % (n = 148), % (n = 148), %
None 26.4 48.7 50.1
Lactobacillus deficient 5.4 10.1 8.1
Gardnerella 5.4 0 0
Bacteroides 0 0 0.7
Gardnerella and Bacteroides 10.8 12.8 14.2
Gardnerella and Lactobacillus deficient 1.4 0 0.7
Bacteroides and Lactobacillus deficient 0.7 0 0
Gardnerella, Bacteroides and Lactobacillus deficient 31.8 16.9 17.6
Gardnerella, Bacteroides, Lactobacillus deficient and Mobiluncus 18.2 11.5 6.8
Modified from Waters et al. (2008)
an important role in preventing infection in genital sites. have found that restoring the flora to normal prevented
The inhibitory effects of lactobacilli against a variety of preterm birth.
microorganisms have long been recognized. Lactobacilli
are more likely to be absent from the vagina in women with Therapeutic options in pregnancy
BV than in women without BV. Therefore, no vaginitis
signs are present in BV. Typically, full-blown uncompli- The question of why some vaginal infection in general, and
cated BV presents the absence of leukocytes on micros- BV in particular, is associated with preterm birth in some
copy, the vagina is not red or inflamed and there are no women but not in others remains still partially unanswered
prominent symptoms of burning, pain or dyspareunia. By and this unavoidably leads to scarce consensus on treat-
contrast, in AV, vaginal leukocytes are usually abundant ment during pregnancy. Given the large amount of evi-
and their numbers as well as their appearance are part of dence that links subclinical maternal infection with preterm
the definition of the disease entity. labor, it has been postulated that the prophylactic use of
Moreover, the microscopic diagnosis of AV is associ- antibiotics in pregnancy or adjunctive use of antibiotics for
ated with a yellow discharge in more than 70% of the preterm labor should result in an improved perinatal out-
women and with vaginal dyspareunia in 12%. come. The results of published studies addressing the issue
The clinical signs of vaginitis, with red inflammation of of antibiotic intervention to prolong gestation vary. In a
the vagina, yellowish discharge and dyspareunia are con- recent review of 15 trials [61], the authors concluded that,
sistent with the microscopy findings of decreased lactoba- despite eradication of BV in pregnancy, antibiotic treat-
cilli, increased vaginal leukocytosis with toxic appearance, ment did not reduce the risk of preterm delivery. There was
parabasal type epithelial cells and increased pH. Further- no prolongation of pregnancy even in women with a
more, the vaginal concentration of succinate, which is history of prior preterm birth. On the other hand, a meta-
mainly produced by anaerobes, was increased in patients analysis of randomized clinical trials showed that macro-
with BV, but not in those with AV. lides and clindamycin given during the second trimester
However, further studies to differentiate the effects of were associated with a lower rate of preterm delivery,
BV and AV on the outcome of pregnancy are urgently whereas metronidazole used alone was linked to a greater
needed. It should also be determined whether BV can risk of preterm delivery in a high-risk population [62].
evolve into AV especially in the third trimester and if Another meta-analysis confirmed that infection screening
this transition can provide insights into the comprehen- and treatment programs in pregnant women may reduce
sion of complications associated with the last phase of preterm birth and preterm low birth weights [63].
pregnancy. Conflicting results in the literature may be attributed to
The pathogenesis of AV and its production of immense variations in study design. Among potential confounders
amounts of vaginal pro-inflammatory cytokines, in fact, are the inclusion of women with preterm contractions
make it an ideal candidate for causing or promoting pre- without infection, administration of antibiotics at varying
term labor, chorioamnionitis and preterm rupture of the stages of intra-uterine infection and at different dosages, or
membranes and could, moreover, help to answer, at least in trials that do not target the appropriate pathogens. For
part, the question why some studies have found no asso- instance, metronidazole is inactive against many organisms
ciation between BV and pregnancy outcome, while others associated with BV and preterm labor, such as M. hominis.
123

10. 598 Arch Gynecol Obstet (2010) 281:589–600
Another possible explanation could be that antibiotics may Some recent evidence associates severe forms of BV in
not effectively treat chorioamnionitis or reverse the pro- pregnancy with poor perinatal outcome, increased risk of
inflammatory mediators that play a key role in the initiation preterm birth with potential neonatal sequelae due to
of labor. Finally, we have to underline that many clinicians prematurity.
do not distinguish between BV and AV. Given that these Vaginal ecosystem study with the detection of patho-
two conditions are responsive to different antibiotic treat- gens is a key instrument in the prevention of preterm
ment (i.e., metronidazole does not have effects on AV, delivery, pPROM, chorioamnionitis, neonatal, puerperal
while clindamycin does), it is difficult to compare their and maternal–fetal infections. The physiological status of
clinical effect to reduce preterm delivery if there is not a the vaginal ecosystem in pregnancy is greatly conditioned
clear characterization of vaginal infection. by Lactobacilli and cell-mediated mechanisms immune of
The ORACLE study [64], which evaluated the use of pregnancy.
broad spectrum, antepartum antibiotics for premature rup- Further decisive clinical trials are necessary to define
ture of fetal membranes, showed some benefits for the baby unanimously the need for correct diagnosis of vaginal
if mothers received erythromycin. Infants of mothers infection, medical treatment during pregnancy and the
treated with erythromycin had less need for oxygen during potential therapeutic protocols.
the hospital stay and a trend toward a reduction in the
composite primary outcome of neonatal death, chronic lung Acknowledgment The authors would like to thank Tracie Dor-
nbusch for revising the manuscript.
disease, or major cerebral abnormality before discharge.
The results from this trial suggest that early antimicrobial Conflict of interest statement None.
treatment could play a role in interrupting the inflammatory
cascade to improve respiratory and other outcomes.
The hormonal changes of pregnancy which favor an
increase in the concentration of lactobacilli might also favor References
the elimination of the BV. Secondly, BV is more common
in sexually active than non-sexually active women. A 1. Witkin SS, Linhares IM, Giraldo P (2007) Bacterial flora of the
female genital tract. function and immune regulation 21(3):347–
reduction in the frequency of sexual intercourses during 354
pregnancy as reported by Read e Klebanoff [65] might, 2. Lidbeck A, Nord CE (1993) Lactobacilli and the normal human
therefore, contribute to a diminishing prevalence of BV. anaerobic microflora. Clin Infect Dis 16(Suppl 4):S181–S187
3. Hawes SE, Hillier SL, Benedetti J, Stevens CE, Koutsky LA,
Wolner-Hanssen P (1996) Hydrogen peroxide-producing Lacto-
bacilli and acquisition of vaginal infections. J Infect Dis
Conclusions 174(5):1058–1063
4. Zhou X, Bent SJ, Schneider MG et al (2004) Characterization of
The typical vaginal flora of a woman during her fertile vaginal microbial communities in adult healthy women using
cultivation-independent methods. Microbiology 150:2565–2573
period is characterized by a remarkable prevalence of 5. Donders GGG (1999) Bacterial vaginosis in pregnancy: screen
Lattobacillus which determines and regulates the physio- and treat? [editorial]. Eur J Obstet Gynecol Reprod Biol 83:1–4
logical acid pH, contributing to the creation and mainte- 6. Verhelst R, Verstraelen H, Claeys G et al (2004) Cloning of 16S
nance of a natural ambient hostile to the attack of microbial rRNA genes amplified from normal and disturbed vaginal
microflora suggests a strong association between Atopobium
pathogens. vaginae, Gardnerella vaginalis and bacterial vaginosis. BMC
Even during gestation, the physiological colonization of Microbiol 4:16–26
the vagina with lactobacilli represents a cornerstone in 7. Fredricks DN, Fiedler TL, e Marrazzo JM (2005) Molecular
preventing infection in genital sites and every effort must identification of bacteria associated with bacterial vaginosis. N
Engl J Med 353:1899–1911
be made to maintain the natural microbial balance in this 8. Rodriguez JM, Collins MD, Sjoden B, Falsen E (1999) Charac-
area. terization of a novel Atopobium isolate from the human vagina:
The natural history of abnormal vaginal flora in preg- description of Atopobium vaginae sp. Nov. Int J Syst Bacteriol
nancy is still poorly understood. 49:e1573–e1576
9. Antonio MA, Hawes SE, Hillier SL (1999) The identification of
It is fundamental to highlight that in pregnancy two vaginal Lactobacillus species and the demographic and micro-
main types of abnormal bacterial flora may occur: anaer- biologic characteristics of women colonized by these species. J
obic and aerobic bacterial vaginitis. There may be a degree Infect Dis 180:e1950–e1956
of interaction and overlap. 10. Stevens-Simon C, Jamison J, McGregor JA, Douglas JM (1994)
Racial variation in vaginal pH among healthy sexually active
The AV with the great production of pro-inflammatory adolescents. Sex Transm Dis 21:e168–e172
cytokines poses as the ideal candidate for the risk of pre- 11. Royce RA, Jackson TP, Thorp JMJ et al (1999) Race/ethnicity,
term birth, preterm premature rupture of membranes vaginal flora patterns, and pH during pregnancy. Sex Transm Dis
(pPROM) and chorioamnionitis. 26:e96–e102
123

11. Arch Gynecol Obstet (2010) 281:589–600 599
12. Fiscella K, Klebanoff MA (2004) Are racial differences in vag- 33. Sobel JD (1994) Desquamative inflammatory vaginitis: a new
inal pH explained by vaginal flora? Am J Obstet Gynecol subgroup of purulent vaginitis responsive to topical 2% clinda-
191:e747–e750 mycin therapy. Am J Obstet Gynecol 171:1215–1220
13. Priestlley CFJ, Jones BM, Dhar J, Goodwin L (1997) What is 34. Donders GGG (1999) Microscopy of bacterial flora on fresh
normal vaginal flora? Genitourin Med 73:e23–e28 vaginal smears. Infect Dis Obstet Gynecol 7:126–127
14. Eschenbach DA, Patton DL, Hooten TM et al (2001) Effects of 35. Gardner HL (1968) Desquamative inflammatory vaginitis: a
vaginal intercourse with and without a condom on vaginal flora newly defined entity. Am J Obstet Gynecol 102:1102–1105
and vaginal epithelium. J Infect Dis 183:e913–e918 36. Donders GGG, Vereecken A, Bosmans E, Dekeersmaecker A,
15. Spiegel CA (1991) Bacterial vaginosis. Clin Microbiol Rev Salembier G, Spitz B (2002) Aerobic vaginitis: abnormal vaginal
4:485–502 flora entity that is distinct from bacterial vaginosis. Br J Obstet
16. Forsum U, Holst E, Larsson PG, Vasquez A, Jacobsson T, Mat- Gynecol 109:1–10
tsby-Baltzer I (2005) Bacterial vaginosis—a microbiological and 37. Mattsby-Baltzer I, Platz-Christensen JJ, Hosseini N, Rosen P
immunological enigma. APMIS 113:81–90 (1998) IL-1beta, IL-6, TNFalfa, fetal fibronectin and endotoxin in
17. Amsel R, Totten PA, Spiegel CA, Chen K, Eshenbach DA, the lower genital tract of pregnant women with bacterial vagin-
Holmes KK (1983) Nonspecific vaginitis. Diagnostic criteria osis. Acta Obstet Gynecol Scand 77:701–706
and microbial and epidemiological associations. Am J Med 38. Imseis HM, Livengood HH, Shunior E, Durda P, Erikson M
74:14–22 (1997) Characterisation of the inflammatory cytokines in the
18. Leopold S (1953) Heretofore undescribed organism isolated from vagina during pregnancy and labour and with bacterial vaginosis.
the genitourinary system. U S Armed Forces Med J 4:263–266 J Soc Gynecol Investig 4:90–92
19. Gardner HL, Dukes CD (1954) New etiologic agent in nonspe- 39. Romero R, Yoon BH, Mazor M, Gomez R et al (1993) The
cific bacterial vaginitis. Science 120:853 diagnostic and prognostic value of amniotic fluid white blood cell
20. Gardner HL, Dukes CD (1955) Haemophilus vaginalis vaginitis: count, glucose, IL-6, and Gram stain in patients with preterm
a newly defined specific infection previously classified non-spe- labour and intact membranes. Am J Obstet Gynecol 169:805–816
cific vaginitis. Am J Obstet Gynecol 69:962–976 40. Chaisilwattana P, Monif GRG (1995) In vitro ability of the group
21. Greenwood JR, Pickett MJ (1979) Salient features of Haemoph- B streptococci to inhibit gram-positive and gram-variable con-
ilus vaginalis. J Clin Microbiol 9:200–204 stituents of the bacterial flora of the female genital tract. Infect
22. Piot P, van Dyck E, Goodfellow M, Falkow S (1980) A taxo- Dis Obstet Gynecol 3:91–97
nomic study of Gardnerella vaginalis (Haemophilus vaginalis) 41. Monif GRG (1999) Semiquantitative bacterial observations with
Gardner and Dukes 1955. J Gen Microbiol 119:373–396 group B streptococci. Infect Dis Obstet Gynecol 7:227–229
23. Swidsinski A, Mendling W, Loening-Baucke V, Ladhoff A, 42. Honig E, Mouton JW, van der Meijden WI (1999) Can group B
Swidsinski S, Hale LP, Lochs H (2005) Adherent biofilms in streptococci cause symptomatic vaginitis? Infect Dis Obstet
bacterial vaginosis. Obstet Gynecol 106:1013–1023 Gynecol 7:206–209
24. Swidsinski A, Mendling W, Loening-Baucke V, Swidsinski S, 43. Usui R, Ohkuchi S, Matsubara S, Izumi A, Watanabe T, Suzuki
¨
Dorffel Y, Scholze J, Lochs H, Verstraelen H (2008) An adherent M (2000) Vaginal lactobacilli and preterm birth. J Perinat Med
Gardnerella vaginalis biofilm persists on the vaginal epithelium 30:458–466
after standard therapy with oral metronidazole. Am J Obstet 44. Waites KB, Katz B, Schelonka RL (2005) Mycoplasmas and
Gynecol 198(97):e1–e6 Ureaplasmas as neonatal pathogens. Clin Microbiol Rev 18:757–
25. Aroutcheva AA, Simoes JA, Behbakht K, Faro S (2001) Gard- 789
nerella vaginalis isolated from patients with bacterial vaginosis 45. Donders GGG (2007) Definition and classification of abnormal
and from patients with healthy vaginal ecosystems. Clin Infect vaginal flora. Best Pract Res Clin Obstet Gynaecol 21:355–373
Dis 33:1022–1027 46. Larsson P-G, Platz-Christensen JJ (1990) Enumeration of clue
26. Waites KB, Katz B, Schelonka RL (2005) Mycoplasmas and cells in rehydrated air-dried vaginal wet smears for the diagnosis
Ureaplasmas as neonatal pathogens. Clin Microbiol Rev 18:757– of bacterial vaginosis. Obstet Gynecol 76:727–730
789 47. Simhan HN, Caritis SN, Krohn MA, Martinez de Teada B,
27. Hay PE, Lamont RF, Taylor-Robinson et al (1994) Abnormal Landers DV, Hillier SL (2003) Decreased cervical proinflam-
bacterial colonisation of the genital tract and subsequent preterm matory cytokines permit subsequent upper genital tract infection
delivery and late miscarriage. Br Med J 308:295–298 during pregnancy. Am J Obstet Gynecol 189:560–567
28. Platz-Christensen J-J, Larsson P-G, Sundstrom E, Wiqvist N 48. Hillier SL, Martius, Krohn M, Kiviat N, Holmes KK, Eschenbach
(1995) Detection of bacterial vaginosis in wet mount, Papanico- DA (1988) A case control study of chorioamnionic infection and
laou stained vaginal smears and in Gram stained smears. Acta histologic chorioamnionitis in prematurity. N Engl J Med
Obstet Gynecol Scand 74:67–70 319:972–978
29. McDonald HM, O’Loughlin JA, Jolley PT et al (1994) Changes 49. Leitch H, Kiss H (2006) asymptomatic bacterial vaginosis and
in vaginal flora during pregnancy and association with preterm intermediate flora as risk factors for adverse pregnancy outcome.
birth. J Infect Dis 170:724–728 Best Pract Res Clin Obstet Gynaecol 21:375–390
30. Spiegel CA, Amsel R, Holmes KK (1983) Diagnosis of bacterial 50. Allsworth JE, Peipert JF (2007) Prevalence of bacterial vaginosis:
vaginosis by direct Gram stain of vaginal fluid. J Clin Microbiol 2001–2004 National Health and Nutrition Examination Survey
18:e170–e177 Data. Obstet Gynecol 109:114–120
31. Kalinka J, Laudanski T, Hanke W, Wasiela M (2003) Do 51. Britton T, Dawn PM (2007) Risk factors for bacterial vaginosis
microbiological factors account for poor pregnancy outcome during pregnancy among African American women. Am J Obstet
among unmarried pregnant women in Poland? Fetal Diagn Ther Gynecol 197:477.e1–477.e8
18:e345–e352 52. Mc Donald H, O’Loughlin JA, Jolley PT, Vigneswaran R,
32. Donders GG, Veerecken A, Bosmans E, Dekersmaecker A, Sa- McDonald P (1994) Changes in vaginal flora during pregnancy
lembier G, Spitz B (2002) Definition of a type of abnormal and association with preterm birth. J Infect Dis 170:724–728
vaginal flora that is distinct from bacterial vaginosis: aerobic 53. Fischbach F, Petersen E, Weissenbacher ER, Martius J, Hossman
vaginitis. Br J Obstet Gynecol 109:34–43 J, Mayer H (1993) Efficacy of clindamycin vaginal cream versus
123

12. 600 Arch Gynecol Obstet (2010) 281:589–600
oral metronidazole I the treatment of bacterial vaginosis. Obstet 67. Edwards RK, Ferguson KJ, Duff P (2006) The interleukin-l beta
Gynecol 82(3):405–410 ?3953 single nucleotide polymorphism: cervical protein con-
54. Waters TP, Denney JM, Mathew L, Goldenberg RL, Culhane JF centration and preterm delivery risk. Am J Reprod Immunol
(2008) Longitudinal trajectory of bacterial vaginosis during 55:259–264
pregnancy. Am J Obstet Gynecol 199:431–435 68. Goffinet F, Maillard F, Mihoubi N et al (2003) Bacterial vagin-
55. Hillier SL, Nugent RP, Eschenbach DA, Krohn MA, Gibbs RS, osis: prevalence and predictive value for premature delivery and
Martin DH (1995) Association between bacterial vaginosis and neonatal infection in women with preterm labour and intact
preterm delivery of a low-birth-weight infant. N Engl J Med membranes. Eur J Obstet Gynecol Reprod Biol 108:146–151
333(26):1737–1742 69. Guerra P, Ghi T, Quarta S et al (2006) Pregnancy outcome alter
56. Krohn MA, Thwin SS, Rabe LK, Brown Z, Hillier SL (1997) early detection of bacterial vaginosis. Eur J Obstet Gynecol
Vaginal colonization by Escherichia coli as a risk factor for very Reprod Biol 128:40–45
low birth weight delivery and other perinatal complications. J 70. Oakeshott P, Kerry S, Hay S, Hay P (2004) Bacterial vaginosis
Infect Dis 175(3):606–610 and preterm birth: a prospective community-based cohort study.
57. Newton ER (1993) Chorioamnionitis and intraamniotic infection. Br J Gen Pract 54:119–122
Clin Obstet Gynecol 36:795–808 71. Andrews WW, Sibai BM, Thom EA et al (2003) Randomized
58. Andrews WW, Goldenberg RL, Hauth C, Cliver SP, Conner M, clinical trial of metronidazole plus erythromycin to prevent
Goepfert AR (2005) Endometrial microbial colonization and spontaneous preterm delivery in fetal fibronectin-positive women.
plasma cell endometritis after spontaneous or indicated preterm Obstet Gynecol 101:847–855
versus term delivery. Am J Obstet Gynecol 193(3 Pt 1):739–745 72. Daskalakis G, Papapanagiotou A, Mesogitis S et al (2006) Bacterial
59. Cassell GH, Davis RO, Waites KB et al (1983) Isolation of vaginosis and group B streptococcal colonization and preterm
Mycoplasma hominis and Ureaplasma urealyticum from amniotic delivery in a low-risk population. Fetal Diagn Ther 21:172–176
fluid at 16–20 weeks of gestation: potential effect on outcome of 73. De Seta F, Sartore A, Piccoli M et al (2005) Bacterial vaginosis
pregnancy. Sex Transm Dis 10(4 Suppl):294–302 and preterm delivery: an open question. J Reprod Med 50:313–
60. Waites KB, Katz B, Schelonka RL (2005) Mycoplasmas and 318
Ureaplasmas as neonatal pathogens. Clin Microbiol Rev 18:757– 74. Genc MR, Witkin SS, Delaney ML et al (2004) A dispropor-
789 tionate increase in IL-1beta over IL-1ra in the cervicovaginal
61. McDonald HM, Brocklehurst P, Gordon A (2007) Antibiotics for secretions of pregnant women with altered vaginal microflora
treating bacterial vaginosis in pregnancy. Cochrane Database correlates with preterm birth. Am J Obstet Gynecol 190:1191–
Syst Rev 1:CD000262 1197
62. Morency AM, Buold E (2007) The effect of second-trimester 75. Goyal R, Sharma P, Kaur I et al (2004) Bacterial vaginosis and
antibiotic therapy on the rate of preterm birth. J Obstet Gynaecol vaginal anaerobes in preterm labour. J Indian Med Assoc
Can 29:35–44 102:548–553
63. Swadpanich U, Lumbiganon P, Prasertcharoensook W, Laopai- 76. Kalinka J, Sobala W, Wasiela M, Brzezinska-Blaszczyk E (2005)
boon M (2008) Antenatal lower genital tract infection screening Decreased proinflammatory cytokines in cervicovaginal fluid, as
and treatment programs for preventing preterm delivery. Coch- measured in midgestation, are associated with preterm delivery.
rane Database Syst Rev 2:CD006178 Am J Reprod Immunol 54:70–76
64. Kenyon SL, Taylor D, Tarnow-Mordi W (2001) Broad-spectrum 77. Kiss H, Petricevic L, Husslein P (2004) Prospective randomised
antibiotics for preterm, prelabour rupture of fetal membranes: the controlled trial of an infection screening programme to reduce the
ORACLE I randomised trial. ORACLE Collaborative Group. rate of preterrn delivery. BMJ 329:371
Lancet 357(9261):979–988 78. Purwar M, Ughade S, Bhagat B et al (2001) Bacterial vaginosis in
65. Read JS, Klebanoff MA (1993) Sexual intercourse during preg- early pregnancy and adverse pregnancy outcome. J Obstet
nancy and preterm delivery: effects of vaginal microorganism. Gynaecol Res 27:175–181
The Vaginal infection and Prematurity Study Group. Am J Obstet 79. Thorsen P, Vogel I, Olsen J et al (2006) Bacterial vaginosis in
Gynecol 168(2):514–519 early pregnancy is associated with low birth weight and small for
66. Donders GGG, Van Bulck B, Caudron J et al (2000) Relationship gestational age, but not with spontaneous preterm birth: a popu-
of bacterial vaginosis and mycoplasmas to the risk of spontaneous lation-based study on Danish women. J Matem Fetal Neonatal
abortion. Am J Obstet Gynecol 183:431–437 Med 19:1–7
123