J Ped Neonatal, 2005; 2(1): NT49-53 WWW. JPN-ONLINE .

ORG
Clinical Neonatology
NT

Received: 2005.01.20
Accepted: 2005.01.21 An association between abnormal vaginal flora during
Published: 2005.03.01
pregnancy and the risk for early-onset neonatal
infection
Authors’ Contribution:
A Study Design Małgorzata Wasiela1 ABDEFG, Paweł Krajewski 2 CDEF, Jarosław Kalinka2 ABCDEFG
B Data Collection
1
C Statistical Analysis Department of Medical Microbiology, Medical University of Łódź, Łódź, Poland
2
D Data Interpretation Department of Perinatology, I Division of Gynaecology and Obstetrics, Medical University of Łódź, Łódź, Poland
E Manuscript Preparation
F Literature Search Source of support: Departmental sources.
G Funds Collection

Summary
Background: The main aim of his study was to determine the relationship between lower genital tract genital
mycoplasmas and BV as diagnosed during pregnancy early-onset neonatal infection.
Material/Methods: The study population comprised 120 pregnant women between 22 and 36 weeks of pregnancy.
The vaginal swabs were testing for diagnosis of Bacterial vaginosis (BV) by Gram’s stain method ac-
cording to Spiegel’s criteria. Cervical swabs were collected for isolation and identification of gen-
ital mycoplasmas. Early-onset neonatal infection was diagnosed on clinical and laboratory symp-
toms of the newborns.
Results: Among 120 examined pregnant women BV was diagnosed in 32 (26.7%)women, 35 (29.1%) had
intermediate flora and among 53 (44.2%) normal flora was diagnosed. Genital mycoplasmas was
diagnosed in 37 subject (30.8%). 21 (17.5%) of the 120 study women delivered newborns with ear-
ly-onset neonatal infection (EONI) (group I), while 99 (82.5%) constituted reference group (non-
IW – group II). Co-infection of BV with mycoplasmas or genital mycoplasmas infection diagnosed
at midgestation constituted an important risk for delivering an infant with early-onset neonatal in-
fection (OR =7.11, and OR =4.44, respectively).
Conclusions: Bacterial vaginosis and the lower genital tract colonization by Mycoplasma hominis and Ureaplasma
urealyticum between 22 and 36 weeks of gestation constitute risk factors for early-onset neonatal
infection what indicate the need for detection and monitoring of this infections among pregnant
women even before pregnancy.

Key words: bacterial vaginosis • genital mycoplasmas • risk factors • early-onset neonatal infection

Full-text PDF: http://www.jpn-online.org/get_pdf.php?IDMAN=6910

Word count: 1822
Tables: 3
Figures: —
References: 36

Author’s address: Dr. n. med. Małgorzata Wasiela, Zakład Mikrobiologii Lekarskiej Uniwersytetu Medycznego w Łodzi,
ul. Pomorska 251, 92-215 Łódź, Poland, e-mail: [email protected]

NT49
Clinical Neonatology J Ped Neonat, 2005; 2(1): NT49-53

BACKGROUND based on medical records. Routine ultrasound examination

of fetal biometry was performed. The gestational age at the
Early-onset neonatal infection are one of the most common time of sampling was based on the time of last menstrua-
perinatal complication. Four possible mechanisms exist for tion and verified by early ultrasound (CRL – crown-rump
microbes to spread to the uterus, an otherwise sterile envi- length) of the fetus.
ronment: (i) the organisms originate from elsewhere and are
transmitted hematogenously; (ii) the organisms from the peri- Bacteriological examination
toneal cavity seed retrogradely through the fallopian tube;
(iii) they are inoculated accidentially inside the uterus dur- For the assessment of biocenosis, vaginal and cervical swabs
ing invasive procedures, such as amniocentesis and chorionic were collected from all pregnant women under study. The
villous sampling, and (iv) the organisms from vagina and the vaginal swabs were testing for Bacterial vaginosis (BV).
cervix ascend to the uterus. Ascending infection is considered Bacterial vaginosis was diagnosed by Gram’s stain accord-
to be by far the most common route of infection [1]. ing to Spiegel’s criteria [11], and the flora were graded as
follows:
Abnormal microbiological flora of the lower genital tract Grade I – normal – predominantly lactobacillus morpho-
– Bacterial vaginosis (BV) is clinical syndrome definied as a types;
decrease in the normally occurring lactobacillus species in Grade II – intermediate – mixed lactobacillus and oth-
the vagina and an increase in other predominantly anaer- er Gram-positive and Gram-negative bacterial
obic organismas, including Gardnerella vaginalis, Bacteroides morphotypes;
spp, Prevotella spp, Mobiluncus spp, Fusobacterium spp, and gen- Grade III – bacterial vaginosis (BV) – few or absent lacto-
ital mycoplasmas – Mycoplasma hominis and/or Ureaplasma bacillus morphotypes, but greatly increased
urealyticum [2,3]. number of G.vaginalis and other anaerobic
Gram-negative roads.
Bacterial vaginosis associated flora are well knows cause
of perinatal complication. It has been documented that Cervical swabs were collected for isolation and identifi-
BV and/or genital mycoplasmas are an important risk fac- cation of genital mycoplasmas (commercially available
tors for chorioamnionitis and amniotic fluid (AF) infec- Mycoplasma DUO kits, Sanofi Diagnostics Pasteur were
tion [4–10]. Although BV associated flora and genital myc- used). Identification was based on specific hydrolysis of
oplasmas have been isolated as a infectious agent from urea (U.urealyticum) or arginine (M.hominis) by the species
amniotic fluid among pregnant women, a casual relation- present in specimen, which is indicated by change in the
ship with colonization of lower genital tract in early preg- colour of the well containing the substrate, without cloud-
nancy and early-onset neonatal infection in newborns are ing in the medium.
still controversial.
Early-onset neonatal infection
The aim of this study was to determine whether lower gen-
ital tract genital mycoplasmas infections and BV, as diag- The early-onset neonatal infection (EONI) defined as infec-
nosed between 22 and 36 weeks of pregnancy, are associat- tions occurring within first 48 hours of birth. The EONI was
ed with early-onset neonatal infection defined as presence of 3 of the following clinical signs sug-
gestive of infection: lethargy or poor feeding; axillary tem-
MATERIAL AND METHODS perature <36ºC or >38º C for more than one hour; signifi-
cant jaundice with serum bilirubin >15 mg% in the absence
Study population of blood group incompatibility; apnoea or respiratory dis-
tress; peripheral capillary refill time of >3 sec on the fore-
The study population comprised 120 women at 22 to 34 head or mid sternum; heart rate of >160/min corrected for
weeks’ gestation recruited from the patients of the clinical elevation of body temperature (10 beats/1ºC rise); vomiting,
hospital at the Department of Perinatoloy, Medical University diarrhoea or ileus; petechiae or bleeding diathesis; omphali-
of Łódź, Poland, between May 2001 and December 2002. tis; seizures and 1 of the following laboratory markers: total
Only the singleton pregnancies were qualified for inclusion leukocyte count <5,000/mm3, neutrophil count <1,500/mm3,
in the survey. The exclusion criteria were as follows: antibi- and immature to total neutrophil ratio >0.2.
otic therapy within 4 weeks prior to the examination, mul-
tiple gestation, vaginal bleeding, serious maternal diseases Statictical analysis
and any immunological disorders. Of 120 women enrolled
six women were lost to follow up and excluded from further For comparison of prevalence of genital mycoplasmas and
analysis so final study group comprised 114 women. BV2 c2 was used. Odds ratios (OR) were calculated to eval-
uate risk factors. Statistically analysis was conducted using
This prospective cohort study was approved by the EPI INFO software, taking into account the odds ratios and
Biomedical Ethics Committee of the Medical University of 95% confidence intervals (CI). Statistical analysis was car-
Łódź, Poland. (Decision No. RNN/215/00). Each partic- ried out using STATA 8 software
ipant provided a written informed consent to participate
in the study. RESULTS
A standard questionnaire covering medical, socio-econom- The mean gestational age at the time of microbiological sm-
ic, demographic and constitutional aspects as well as tobac- plings was 29.0 (SD 4.0) weeks. The mean maternal age of
co smoking was administered to every subject and verified the study group was 27.3 (SD 4.7) years. 21% of the subject

NT50
J Ped Neonat, 2005; 2(1): NT49-53 Wasiela M et al – Abnormal vaginal flora and early-onset infection

Table 1. The prevalence of BV and genital mycoplasmas among women who delivered infants with early-onset neonatal infection (EONI)
and among women who delivered infants without early-onset neonatal infection (non EONI).
NT
Gram-stain mycoplasmas EONI (n=21) non EONI (n=99) p
BV 9 42.9% 23 23.2% ns
Intermediate flora 5 23.8% 30 30.3% ns
Normal flora 7 33.3% 46 46.5% ns
mycoplasmas 13 61.9% 24 24.2% p<0.01

Table 2. The relation in prevalence of BV and genital mycoplasmas.

Gram-stain mycoplasmas EONI (n=21) non EONI (n=99) p

BV 2 9.5% 11 11.1% ns
BV and mycoplasmas 8 38.0% 12 12.1% p<0.01
Mycoplasmas and intermediate flora 2 9.5% 5 5.5% ns
Mycoplasmas and normal flora 3 14.3% 7 7.1% ns
Intermediate and normal flora 6 28.6% 64 64.6% p<0.01

Table 3. Association between BV and genital mycoplasmas diagnosed at 22 weeks gestation and the risk of delivering infant with early-onset
neonatal infection.

EONI
+ – OR 95%CI
(n=21) (n=99)
BV(–) mycoplasmas (–) 6 28.6% 64 64.6% Reference group
BV(–) mycoplasmas (+) 5 23.8% 12 10.0% 4.44 (0.90–20.42)
BV(+) mycoplasmas (–) 2 9.52% 11 11.1% 1.94 (0.17–12.72)
BV(+) mycoplasmas (+) 8 38.1% 12 12.1% 7.11 (1.81–29.05)

were unmarried and 16.1% had primary education. In the 9 (42.9%) of women of group I had BV and 13 (61.9%) were
study population, 60.5% of women were nulliparous and culture-positive for genital mycoplasmas as compared to 23
15.8% were smoking during pregnancy. (23.2%) and 24 (24.2%, p=0.033) in the group II (Table 1).

Microbiological and clinical characteristic We also evaluated the relation in prevalence of BV and gen-
ital mycoplasmas between women from group I and group
Among 120 examined pregnant women BV was diagnosed II. BV was diagnosed alone in 2 (9.5%) cases from group I
in 32 (26.7%) women, 35 (29.1%) had intermediate flora and in 11 (11.1%) cases from group II. The most frequent-
and among 53 (44.2%) normal flora was diagnosed. Genital ly co-infection BV/mycoplasmas were diagnosed in group
mycoplasmas was diagnosed in 37 subject (30.8%). I – 8 (38.1) as compared to 12 women (12.1%) from the
group II (Table 2).
21 (17.5%) of the 120 study women delivered newborns
with early-onset neonatal infection (EONI) (group I), Co-infection BV/mycoplasmas or genital mycoplasmas infec-
while 99 (82.5%) constituted reference group (non-EONI tion diagnosed during pregnancy constituted an important
– group II). risk for delivering an infant with early-onset neonatal infec-
tion (OR =7.11, and OR =4.44; respectively) (Table 3).
Microbiological results and early-onset neonatal infection
DISCUSSION
Women who delivered infants with EONI were more likely
to be BV and genital mycoplasmas positive (both M.hominis The prevalence of bacterial vaginosis among pregnant wom-
and U.urealyticum or alone) than those in references group. en ranges from 12% to 50%, depending on the population

NT51
Clinical Neonatology J Ped Neonat, 2005; 2(1): NT49-53

studied [12–14]. In our study BV was diagnosed among been demonstrated that some microbes are able to cross
26.7% of pregnant women at 22–36 weeks of gestation by the intact membranes, however, asymptomatic women with
Gram’s stain using Spiegel’s criteria. The relatively high in- intact membranes may also sometimes (incidence to 24%)
cidence of BV could be explained by use of sensitive diag- have bacteria in their amniotic fluid [32,33]. In previous
nostic method. In clinical practice bacterial vaginosis is usu- study we also documented that bacterial vaginosis and gen-
ally diagnosedon the basis of composite Amsel’s criteria. ital mycoplasmas in early pregnancy are the risk factors for
According to Tam et al. [15] sensitivity of Gram stain meth- low birth weight and preterm delivery [34–36].
od was significantly higher than that of clinical criteria (91%
vs 46%). The Gram stain method has both a low false-nega- The results of this study suggest that bacterial vaginosis and
tive (4%) and high negative predictive values (96%). lower genital tract colonization by Mycoplasma hominis and
Ureaplasma urealyticum, as diagnosed at midgestation consti-
It was shown, that Mycoplasma hominis and Ureaplasma tute risk factors for early-onset neonatal infection what in-
urealyticum are the most common organisms isolated in the dicate the need for detectable and monitoring of pregnant
perinatal period and are associated with the more frequent women even before pregnancy for thease abnormal bacte-
occurrence of bacterial vaginosis [16–19]. In our study gen- rial flora. This could indicate the need for detection and
ital mycoplasmas were isolated among 30% of the 120 preg- monitoring of this infections among pregnant women ear-
nant women and among 62.5% of women with BV. ly at or even before pregnancy.

Bacterial vaginosis associated flora and genital mycoplasmas CONCLUSIONS

are well knows causes of intraamniotic infection. A variety of
microorganisms have been isolated from the placenta, am- Bacterial vaginosis and the lower genital tract colonization
niotic fluid and chorioamnion cultures [20–22]. The most by Mycoplasma hominis and Ureaplasma urealyticum be-
common microorganisms involved in intrauterine infections tween 22 and 36 weeks of gestation constitute risk factors
are Ureaplasma urealyticum, Mycoplasma hominis and other bac- for early-onset neonatal infection what indicate the need
terial vaginosis associated flora, especially Fusobacterium and for detection and monitoring of this infections among preg-
Bacteroides species. Hitti et al. [23] isolated the same microor- nant women even before pregnancy.
ganisms from amniotic fluid (AF) culture and vagina. 72%
of the women with positive AF culture had bacterial vagi- REFERENCES
nosis-associated flora isolated from AF. Amniotic fluid in-
fection was associated with absence of hydrogen peroxide- 1. Romero R, Avila C, Brekus A, Morotti R: The role of systemic
producing Lactobacillus, and presence of vaginal Bacteroides and intrauterine infection in preterm parturition. Ann N Y Acad Sci,
1991; 622: 255–75
urealyticus and Fusobacterium spp. Jacobson et al. [24] inves-
2. McGregor JA, French JI: Bacterial vaginosis in pregnancy. Obstet
tigated the occurrence of amniotic fluid infection caused Gynecol Surv, 2000; 55 (Suppl.1): 1–19
by genital M.hominis and U.urealyticum and anaerobic bac- 3. Mendoza-Gonzales A, Sanchez-Vega JT, Sanches-Peon I et al:
teria in population of Swedish women with preterm prela- Frequency of Gardnerella vaginalis vaginosis and its association with oth-
bor rupture of membranes. Microorganisms in amniotic er pathogens causing genital infections in the female. Ginecol Obstet
fluid were detected in 25% patients. Rzanek-Głowacka et Mex, 2001; 69: 272–76
al. [25] evaluated the relationship between bacterial flora 4. Gibbs RS, Romero R, Hillier SL et al: A review of premature birth
and subclinical infection. Am J Obstet Gynecol, 1992; 166: 1515–28
taken from cervix of pregnant women with premature rup-
5. Colli E, Bertulessi C, Landoni M et al: Bacterial vaginosis in
ture of membranes (PROM) and presence of clinical and pregnancy and preterm birth: evidence from literature. J Int Med Res,
labolatory symptoms of infections of the newborn. On the 1996; 24: 317–24
basis on analyzed values PROM among mothers with bacte- 6. Paul VK, Grupta U, Singh M et al: Association of genital myco-
rial vaginosis was the important risk factor of the early-on- plasma colonization with low birth weight. Int J Gynaecol Obstet, 1998;
63: 109–14
set neonatal infection.
7. Egger M, Mmuhlemann K, Aebi C, Tauber MG: Infections in
pregnancy. Ther umsech, 1999; 56: 577–82
In our study bacterial vaginosis and genital mycoplasmas 8. Georgijevic A, Cjukic-Ivancevic S, Bujko M: Bacterial vaginosis.
have been associated with vaginal colonization in women Epidemiology and risk factors. Srp Arh Celok Lek, 2000; 128: 29–33
at 22–36 weeks’ gestation and constituted a significant risk 9. McGregor JA, French JI: Bcterial vaginosis in pregnancy. Obstet
factors for delivering infant with early-onset neonatal in- Gynecol Surv, 2000; 55 (5 Suppl 1): 1–19
fection (OR = 7.11 CI: 1.81–29.05). BV was detected in al- 10. Calleri LF, Taccani C, Porcelli A: Ureaplasma urealyticum vaginosis and
most half of the women (42.9%) and genital mycoplasmas premature rupture of membranes. What is its role? Minerva Ginecol,
2000; 52: 49–58
among 57.4% women who delivered infants with early-on-
11. Spiegel CA, Amsel R, Holmes KK: Diagnosis of bacterial vaginosis by di-
set neonatal infection. Povlsen et al. [26] have reported that rect Gram-stain of vaginal fluid. J Clin Microbiol, 1983; 18: , 170–77
U. urealyticum may be the cause infection independently of 12. McGregor JA, French JI: Bacterial vaginosis in pregnancy. Obstet Gynecol
bacterial vaginosis. We also detected BV in 9.52% of wom- Surv, 2000; 55(Suppl.1): 1–19
en without mycoplasmas infection and M.hominis and/or 13. Mendoza-Gonzales A, Sanchez-Vega JT, Sanches-Peon I et al: Frequency
U. urealyticum without sings of BV. of Gardnerella vaginalis vaginosis and its association with other path-
ogens causing genital infections in the female. Ginecol Obstet Mex,
2001; 69: 272–76
Many studies have shown that early-onset neonatal infection
14. Wilson J: Managing recurrent bacterial vaginosis. Sex Transm Infect,
are highly prevalent among women who have preterm la- 2004; 80: 8–11
bor. The presence of microorganisms in amniotic fluid usu- 15. Tam MT, Yungbluth M, Myles T: Gram stain method shows better sen-
ally results in spontaneous abortion, prematurity and fetal sitivity than clinical criteria for detection of bacterial vaginosis in sur-
or neonatal infections. AF infection occurs in 10–15% of veillance of pregnant, low-income women in clinical setting. Infect Dis
Obstet Gynecol, 1998; 6: 204–9
pregnancies complicated by preterm labor [27–31]. It has

NT52
J Ped Neonat, 2005; 2(1): NT49-53 Wasiela M et al – Abnormal vaginal flora and early-onset infection

16. Spiegel CA: Bacterial vaginosis. Clin Microbiol Rev, 1991; 4: 485–502
17. Gardo S: Bacterial vaginosis. Orv. Hetil, 1998; 139: 1403–8
27. Watts DH, Krohn MA, Hillier SL, Eschenbach DA: The association of oc-
cult amniotic fluid infection with gestation age and neonatal outcome
among women in preterm labor. Obstet Gynecol, 1992; 79: 351–57
NT
18. McGregor JA, French JI: Bcterial vaginosis in pregnancy. Obstet Gynecol
Surv, 2000; 55(5 Suppl 1): 1–19 28. Chaim W, Mazor M, Leiberman JR: The relationship between bacterial
vaginosis and preterm birth. A revive. Arch Gynecol Obstet, 1997; 259:
19. Povlsen K, Thorsen P, Lind I: Relationship of Ureaplasma urealyticum 51–58
Biovars to the presence or absence of bacterial vaginosis in pregnant
women and to the time of delivery. Eur J Clin Microbiol Infect Dis, 2001; 29. Krohn MA, Hitti J: Characteristics of women with clinical intraamniot-
20: 65–67 ic infection who deliver preterm compared with term. Am j Epidemiol,
1998; 147: 111–16
20. Chaim W, Mazor M: Iintraamniotic infection with fusobacteria. Arch
gynecol Obstet, 1992; 251: 1–7 30. Hill GB: Preterm birth: association with genital and possibly oral mi-
croflora. Ann Peridontol, 1998; 3: 222–32
21. Hill GB: Investigating the source of amniotic fluid isolates of fusobac-
teria. Clin Infect Dis, 1993; 16(Suppl.4): S423–S424 31. Goncalves LF, Chaiworapongsa T, Romero R: Intrauterine infection
and prematurity. Ment Retard Dev Disabil Res Rev, 2002; 8: 3–13
22. Yoon BH, Romero R, Kim M et al: Clinical implication of detection of
Ureaplasma urealyticum in the amniotic cavity with the polymerase 32. Dunlow SG, Duff P: Microbiology of the lower genital tract and amni-
chain reaction. Am J Obstet, Gynecol, 2000; 183: 1130–37 otic fluid in asymptomatic preterm patients with intact membranes and
moderate to advanced degrees of cervical effacement and dilatation.
23. Hitti J, Hillier SL, Agnew KJ et al: Vaginal indicators of amniotic fluid Am J Perinatol, 1990; 7: 235–38
infection in preterm labor. Obstet gynecol, 2001; 97: 211–19
33. Mandar R, Loivukene K, Ehrenberg A et al: Amniotic fluid microflora
24. Jacobson B, Mattsby-Balzer I, Andersch B et al: microbial invasion and in asymptomatic women at mid-gestation. Scand J Infect Dis, 2001; 33:
cytokine response in amniotic fluid in Swedish population of women 60–62
with preterm prelabor rupture of membranes. Acta Obstet Gynecol
Scand, 2003; 82: 423–31 34. Kalinka J, Laudański T, Hanke W, Wasiela M: Prevalence and the im-
pact of bacterial vaginosis and bacterial vaginosis-associated flora at
25. Rzanek-Głowacka J, Pieta-Dolińska A, Zięba K, Oszukowski P: Is the early pregnancy on te risk of low birth weight. Archives of Perinatal
mother’s bacterial vaginosis with PROM a significant factor for intrau- Medicine, 2003; 9: 34–37
terine infection of the fetus in preterm labor before 32 weeks of gesta-
tion. Ginekol pol, 2003; 74: 1262–68 35. Wasiela M, Krzemiński Z, Hanke W, Kalinka J: An association between
genital mycoplasmas and risk of preterm delivery. Medycyna Wieku
26. Povlsen K, Thorsen P, Lind I: Relationship of Ureaplasma urealyticum Rozwojowego, 2003; 3(Suppl.1): 211–16
biovars to the presence or absence of bacterial vaginosis in pregnant
women and to the time of delivery. Eur J Clin Microbiol Infect Dis, 2001; 36. Kalinka J, Laudański T, Hanke W et al: The evaluation of prevalence and
20: 65–67 the impact of pathological microflora of the lower genital tract among
women at early pregnancy on the risk of preterm delivery. Med Dośw
Mikrobiol, 2003; 55: 277–84

NT53