Discussion
Main Findings
There were three important findings in this study: 1) Almost a third of
all women delivering term or near neonates were colonized with GBS in
this hospital; 2) even after controlling for various confounders,
smoking during pregnancy was more than twice more likely to result in
GBS colonization among pregnant women of term or near term neonates.; 3)
teen mothers less than 20 years had the highest rates of GBS
colonization and the rates steadily decreased with age with the lowest
rates occurring among the oldest mothers 35 year+. Interestingly this
inverse relationship was only true for women who were nonsmokers.
Interpretations
This high prevalence of GBS colonization found in our study is
consistent with previous studies in the USA and
elsewhere.3-11 The lowest rates were observed mainly
in some East Asian nations.5,11,12 The across-country
differences may be due to differences in
techniques.23-25 PCR testing after broth enrichment
has been has been shown to be a reliable and is now the standard method
for GBS testing.24 However, testing conducted
intrapartum may show somewhat lower prevalence rates as compared to
antepartum testing.25 If this is so, then there may be
excessive use of intrapartum prophylactic antibiotics to prevent GBS
sepsis in neonates. Unfortunately, intrapartum testing is not yet
practical since it can take up to 2 to 3 days for the results of tests
to be made available whereas knowing the results prior to delivery is
very important to make a decision to treat prophylactically or not.
Hopefully, new rapid PCR tests are being developed for such a
purpose.25 In spite of the dramatic drop of early
onset neonatal GBS sepsis in the United States, the reasons for the
persisting high prevalence rates of GBS colonization during pregnancy
throughout the past 4 decades need to be further explored.
This study further confirms that, potentially, more than a third of all
women were eligible to receive intrapartum prophylactic antibiotics
prior to delivery. This is indeed a high rate of pre-delivery
antibiotics exposure of term or near term neonates. Several authors have
recently suggested that at least 40% of children may be exposed to
intrapartum antibiotics.26,27 Although intrapartum
antibiotic prophylaxis have successfully reduced the incidence of GBS
sepsis in neonates, the dangers of such prophylaxis are
concerning.28 For instance, intrapartum antibiotic
prophylaxis have been associated with complications such as increased
prevalence of atopic dermatitis,29 antimicrobial
resistance,30 and changes in the neonatal gut
microbiota.31-33
The prevalence of active tobacco smoking among pregnant women in this
study was 12%, and this is consistent with recent national
trends.22 However, in our study, the intensity of
smoking during pregnancy was rather low and most women who smoked
admitted to smoking no more than 3 cigarettes each day although this may
be largely an underestimate. Since this was a retrospective study the
assumption is that women who admitted to smoking actually smoked
throughout pregnancy. The study may, therefore, not take account of
those who stopped smoking sometime during pregnancy or those who only
initiated smoking sometime during pregnancy.
Few studies in the literature have explored the impact of smoking during
pregnancy on GBS colonization. Our study was also able to demonstrate a
significant dose-response relationship suggesting that this was not just
a spurious finding. In the United States, Terry et
al.34 were the first to show that smoking during
pregnancy was predicted of GBS colonization but the authors did not
perform a multivariable analysis to control for other confounders to
determine if smoking was independently predictive. Edwards and
colleagues, 35 in a large retrospective study,
recently demonstrated that smoking was also predictive of GBS
colonization in the in the univariate analysis but not in the multiple
regression analysis. However, smoking was not the primary independent
variable of their study. Two studies respectively from Korea and
China,12, 36 failed to demonstrate
an association of smoking during pregnancy and GBS colonization. However
smoking was also not the primary independent variable of interest of
these studies and there were was the additional problem of smallness of
numbers in some cells. Our finding that smoking during pregnancy was
predictive of GBS colonization during pregnancy is consistent with a
smaller study from Iran. 37 However, their study
population was different as it also included preterm neonates that were
excluded in our current study. Surprisingly, Regan and colleagues
demonstrated in 1991 that smoking was rather protective of GBS
colonization.38 The authors offered no biological
plausibility of their finding. However, only women delivering preterm
neonates (<36 weeks) were enrolled in the latter study whereas
our study focused only on term or near term neonates. We speculate that
the population of women delivering term babies may be quite different
from the population delivering only preterm babies.
The finding that tobacco smoke exposure during pregnancy is an
independent risk factor for increased GBS colonization has significant
public health implications. Because smoking exposure is a modifiable
risk factor, women can be counseled to stop smoking during pregnancy in
order to reduce colonization with this organism which can result in GBS
sepsis in their newborn baby. The association between tobacco smoking
and increased GBS colonization during pregnancy is biologically
plausible. We speculate that tobacco smoke exposure during pregnancy may
actually enhance the colonization of GBS in the gastrointestinal and the
genital tracts. Indeed, previous studies have shown that tobacco smoke
exposure is associated with increased colonization of the respiratory
and genital tract with pathogenic bacteria,39-41possibly through alteration of the microbiome.17,19,41Tobacco smoke contains more than 4,500 chemical
intoxicants42 many of which can result in increased
suppression or modulation of both active and passive immune
response.43 For instance, nicotine, an important
component of tobacco smoke, has been shown to enhance the adherence of
bacteria in mucous membranes of the respiratory tract leading to easy
penetration of bacteria into the tissues to cause
infections.44 This could also be true of the genital
tract of pregnant women where nicotine could actually result in the
persistence of GBS in the mucous membranes of the gastrointestinal and
genital tracts. Indeed some studies have actually demonstrated higher
nicotine levels in the cervical mucous membranes of smokers as compared
to non-smokers.45,46 In another study, nicotine of the
cervical mucus of female smokers resulted in DNA damage of epithelial
cells of these women resulting in easy penetration of bacteria into the
adjacent tissues.47 It can be speculated that the
overall effect of tobacco smoke is the alteration of the microbiome of
the female genital tract resulting in increased prevalence of pathogenic
microorganisms such as GBS in the present study.
To our knowledge only a few studies in the United States have been
conducted in recent years to determine potential sociodemographic risk
factors associated with this rather high colonization rates among
pregnant women. In the ‘70’s Anthony et al showed that Mexican Americans
had the lowest rates as compared to Whites or
Blacks.48 In our study, black mothers had the highest
rates as compared to White or Hispanic/Latino mothers. These findings
are consistent with those of Regan and colleagues whose study was
conducted almost 30 years ago.38 The finding that
younger mothers had significantly higher rates of GBS colonization was
consistent with one previous study conducted more than 40 years ago by
Anthony and colleagues.48 They demonstrated that
younger mothers had higher rates of GBS colonization than their older
counterparts. Indeed, GBS sepsis in neonates has also been shown to be
more common among young mothers < 20 years, as demonstrated by
Schuchat et al.49,50 This may be explained by the fact
that young mothers also have higher rates of GBS colonization than their
older counterparts as shown in our study. However, 10 years after the
study by Anthony et al,48 Regan et al38 showed that GBS colonization was less common among
women less than 20 years of age even after controlling for the other
sociodemographic confounders. The difference in findings between the two
studies may be due to the fact that cultures for GBS were obtained very
early in gestation (23-26 weeks gestation) in the latter study, whereas
the current recommendation is to obtain cultures at 35-37 weeks of
gestation.1Again as stated above, our study clearly
demonstrates that maternal age was only predictive of GBS colonization
among the non-smoking women. Non-smokers < 20 years had the
highest GBS colonization rate while those >35 years of age
had the lowest rate (P of Trend =0.02; also see Fig 1).
Limitations and Significance
This study has several limitations. First the retrospective nature of
the data implies that there is no causality attributed to the findings.
Second, because the study only involved subjects recruited from one
local hospital, there may therefore be lack of both internal and
external validity of the findings as this sample was not necessarily
representative of the population of the state or of the nation. Third,
the information of tobacco smoking was retrospectively obtained so there
is likelihood of misclassification bias. The association between tobacco
smoking and GBS colonization would have been even stronger if the
smoking status was determined objectively by the use of a biomarker such
as serum or urine cotinine levels. This may have reduced the likelihood
of misclassification bias of the smoking status of the subjects.
However, some of the findings are consistent with previous works.
Fourthly, we were not able to control for all the confounders such as
obesity and the frequency of sexual relationships which have been shown
in some studies to be predictive of GBS
colonization.51 The significance of this study lies in
the fact that this is the first robust study with the main focus of
determining the impact of tobacco smoke on GBS colonization on women of
term or near neonates. In other studies GBS was not the main focus and a
test of trend was not explored.