ArticlePDF Available

Can known risk factors explain racial differences in the occurrence of bacterial vaginosis?

Authors:

Abstract and Figures

Black women are more likely to have bacterial vaginosis (BV) than are non-Hispanic white women. We examined whether this disparity can be explained by racial differences in known BV risk factors. Nine hundred black and 235 white women were enrolled from five US sites. At baseline, structured interviews were conducted and vaginal swabs self-collected for Gram-stain and culture. Black women were more likely than white women to have BV/intermediate vaginal flora. They also were more likely to be older, have lower educational attainment and family incomes, have a history of a sexually transmitted disease, and douche. After adjustment for demographic and lifestyle factors, blacks remained at elevated risk for BV/intermediate flora (OR 2.2, 95% CI 1.5-3.1). Blacks also were more likely to have specific BV-related vaginal microflora, as well as gonococcal or chlamydial cervicitis (OR 2.2, 95% CI 1.2-3.8) after adjustment for known BV risk factors. Risk factor differences did not explain the observed racial disparity in the occurrence of BV, BV-related microflora, or gonococcal or chlamydial cervicitis. These findings highlight our limited understanding of the factors accounting for the occurrence of bacterial vaginosis and cervicitis among black and white women.
Content may be subject to copyright.
CAN
KNOWN
RISK
FACTORS
EXPLAIN
RACIAL
DIFFERENCES
IN
THE
OCCURRENCE
OF
BACTERIAL
VAGINOSIS?
Roberta
B.
Ness
MD,
MPH,
Sharon
Hillier,
Holly
E.
Richter,
PhD,
MD,
David
E.
Soper,
MD,
Carol
Stamm,
MD,
Debra
C.
Bass,
MS,
Richard
L.
Sweet,
MD,
and
Peter
Rice
Pittsburgh,
Pennsylvania;
Birmingham,
Alabama;
Charleston,
South
Carolina;
Denver,
Colorado;
Boston,
Massachusetts
Background:
Black
women
are
more
likely
to
have
bacterial
vaginosis
(BV)
than
are
non-
Hispanic
white
women.
We
examined
whether
this
disparity
can
be
explained
by
racial
differences
in
known
BV
risk
factors.
Methods:
Nine
hundred
black
and
235
white
women
were
enrolled
from
five
US
sites.
At
base-
line,
structured
interviews
were
conducted
and
vaginal
swabs
self-collected
for
Gram-stain
and
cul-
ture.
Results:
Black
women
were
more
likely
than
white
women
to
have
BV/intermediate
vaginal
flora.
They
also
were
more
likely
to
be
older,
have
lower
educational
attainment
and
family
incomes,
have
a
history
of
a
sexually
transmitted
disease,
and
douche.
After
adjustment
for
demo-
graphic
and
lifestyle
factors,
blacks
remained
at
elevated
risk
for
BV/intermediate
flora
(OR
2.2,
95%
Cl
1.5-3.1).
Blacks
also
were
more
likely
to
have
specific
BV-related
vaginal
microflora,
as
well
as
gonococcal
or
chlamydial
cervicitis
(OR
2.2,
95%
Cl
1.2-3.8)
after
adjustment
for
known
BV
risk
factors.
Conclusion:
Risk
factor
differences
did
not
explain
the
observed
racial
disparity
in
the
occur-
rence
of
BV,
BV-related
microflora,
or
gonococcal
or
chlamydial
cervicitis.
These
findings
highlight
our
limited
understanding
of
the
factors
accounting
for
the
occurrence
of
bacterial
vaginosis
and
cervicitis
among
black
and
white
women.
(J
Natl
MedAssoc.
2003;95:201-212.)
02003.
From
the
University
of
Pittsburgh
and
Magee
Womens
Hospital,
Pittsburgh,
Pennsylvania;
University
of
Alabama
School
of
Medicine,
Birmingham,
Alabama;
Medical
University
of
South
Carolina,
Charleston,
South
Carolina;
Denver
Health
Medical
Center,
Denver,
Colorado;
Boston
Medical
Center,
Maxwell
Finland
Laboratory,
Boston,
Massachusetts.
Address
correspondence
to:
Roberta
B.
Ness,
University
of
Pittsburgh,
Graduate
School
of
Public
Health,
Room
517
Parran
Hall,
130
DeSoto
Street,
Pittsburgh,
PA
15261;
phone
(412)
624-3045;
fax
(412)
624-1056;
or
send
e-mail
to
repro@pitt.edu.
Key
words:
bacterial
vaginosis
+
BV
*
cervicitis
*
gynecology
Growing
literature
suggests
that
black
women
are
more
likely
than
non-Hispanic
white
women
to
have
clinical
and
microbiologic
evi-
dence
of
bacterial
vaginosis
(BV),
a
condition
in
which
the
normally
protective
hydrogen
perox-
ide
producing
(H202+)
lactobacilli
are
over-
JOURNAL
OF
THE
NATIONAL
MEDICAL
ASSOCIATION
VOL.
95,
NO.
3,
MARCH
2003
201
grown
by
disease-promoting
endogenous
anaer-
obic
and
facultative
aerobic
bacteria,
including
Gardnerella
vaginalis,
Mycoplasma
hominis,
Mobiluncus
species,
and
anaerobic
Gram
nega-
tive
rods.1
Reported
rates
of
BV
range
from
9%
to
57%
among
reproductive
age
women
with
the
variability
explained
by
sexual
patterns,
BV
def-
inition,
and
race.
2-5
The
association
between
black
race
and
BV
has
been
demonstrated
among
both
pregnant
and
non-pregnant
women
in
the
US
and
non-pregnant
women
in
the
UK.2-
5
Caution
in
the
interpretation
of
this
link
is
warranted,
however,
since
a
number
of
sociode-
mographic
and
lifestyle
factors
increase
the
risk
of
BV
and
are
more
common
among
black
women.
These
sociodemographic
and
lifestyle
factors,
which
include
douching,
low
income
and
low
educational
attainment,
number
of
sex-
ual
partners,
a
history
of
sexually
transmitted
infection
including
PID,
previous
pregnancies,
and
lack
of
hormonal
contraceptive
use,
may
produce
the
appearance
of
a
relationship
between
race
and
BV
when,
indeed,
such
a
rela-
tionship
does
not,
in
truth,
exist.3,6,7
Since
race
is
a
social
construct,8'9
it
is
rational
to
suggest
that
relevant
lifestyle,
sociodemographics
and
healthcare
factors
could
explain
racial
differ-
ences
in
the
occurrence
of
BV.
A
handful
of
previous
reports
have
assessed
whether
race
is
a
predictor
for
BV,
independent
of
other
risk
factors,
with
divergent
results.
One
cross-sectional
study
and
a
prospective
cohort
study,
both
ascertaining
women
from
STI
clin-
ics,
did
not
find
race
to
be
an
independent
pre-
dictor,4'10
whereas
three
cross-sectional
studies,
drawing
on
more
diverse
populations,
two
con-
ducted
in
pregnant
women2'11
and
one
in
non-
pregnant
women,
did
find
race
to
predict
BV
status.3
We
report
here,
among
black
and
non-
Hispanic
white
women
participating
in
the
base-
line
examination
of
the
GYN
Infections
Follow-
Through
(GIFT)
Study,
the
associations
between
known
risk
and
protective
factors
for
BV.
We
then
assess
the
degree
to
which
known
risk
fac-
tors
for
BV
might
explain
racial
differences
in
vaginal
microecology.
METHODS
Patient
selection
Between
May
1999,
and
June
2001,
women
13
to
36
years
of
age
were
recruited
from
five
sites
located
in
the
eastern,
southern,
and
west-
ern
regions
of
the
US
into
the
GYN
Infections
Follow-through
(GIFT)
Study,
a
cohort
investi-
gation
of
the
health
consequences
of
douching.
Human
subjects
approval
was
obtained
at
each
participating
institution
and
all
women
signed
informed
consent.
Women
enrolled
from
public
and
private
gynecology
clinics,
student
health
services,
and
health
departments,
were
eligible
for
the
GIFT
study
if
they
were
not
specifically
seeking
care
for
an
STI,
yet
were
at
elevated
risk
based
on
a
previous
risk
stratification
paradigm
for
chlamydial
cervicitis,
as
described
below.12
To
be
eligible,
a
woman
had
to
have
a
score
of
three
points
or
more
on
an
algorithm
wherein
points
were
derived
as
follows:
age
24
or
less
=
1;
black
race
=
2;
never
pregnant
=
1;
two
or
more
sexual
partners
=
1;
douched
at
least
once
per
month
=
2;
any
sexually
transmitted
infec-
tion,
including
N.
gonorrhoeae,
C.
trachomatis,
and
TJ
vaginalis
=
2.
Of
2740
women
screened
for
study
entry,
853
(31.1%)
did
not
meet
these
inclusion
criteria.
An
additional
259
(9.5%)
women
were
excluded
on
the basis
of
a
priori
criteria
including:
currently
pregnant
by
beta
HCG
testing,
currently
married,
never
having
had
sexual
intercourse,
having
had
a
prior
hys-
terectomy,
salpingectomy,
or
tubal
ligation,
or
being
on
antibiotics
at
baseline
(because
of
their
lower
sexually
transmitted
disease
risk);12
z
having
pelvic
tenderness
on
examination
at
base-
line.
Among
the
1628
women
who
were
eligible
for
the
study,
1200
(73.7%)
agreed
to
study
par-
ticipation,
1135
of
whom
identified
themselves
as
black
or
non-Hispanic
white
and
are
the
focus
of
these
analyses.
202
JOURNAL
OF
THE
NATIONAL
MEDICAL
ASSOCIATION
VOL.
95,
NO.
3,
MARCH
2003
Table
1.
NUMBER
(%)
OF
BLACK
AND
WHITE
WOMEN
WITH
SELECTED
BACTERIAL
VAGINOSIS
RISK
FACTORS
Blacks
Whites
p
N
%
N
%
Age
25+
286
31.8
28
11.9
19-24
554
61.6
190
80.9
13-18
60
6.7
17
7.2
p=
.000
Education
>
HS
graduate
436
48.4
159
67.7
HS
graduate
277
30.8
41
17.4
<
HS
graduate
187
20.8
35
14.9
p=
.000
Marital
Status
Living
as
married
46
5.1
7
3.0
Never
married
809
89.9
216
92.3
p=
.379
Divorced/widow/other
45
5.0
11
4.7
Income
3$20,000
154
18.8
120
56.6
$10,000-19,000238
29.0
33
15.6
<
$10,000
428
52.2
59
27.8
p=
.000
History
of:
PID
136
15.2
19
8.1
p=
.005
Chlamydia
390
43.6
50
21.5
p=
.000
BV
355
40.2
52
22.2
p=
.000
Gonorrhea
227
25.4
19
8.2
p=
.000
Trichomoniasis
254
28.5
26
11.1
p=
.000
Gravidity
Ever
pregnant
582
64.7
52
21.1
p=
.000
Tobacco
smoking
Never
550
61.2
80
34.0
p=.000
Former
66
7.3
41
17.4
Current
282
31.4
114
48.5
Number
of
Partners
(past
2
months)
None
138
15.3
35
14.9
p=
.001
One
639
71.0
144
61.3
3
2
123
13.7
56
23.8
New
partners
(past
2
months)126
16.5
70
35.0
p=
.000
Anal
sex
(past
2
months)
47
5.2
19
8.1
p=
.095
Sex
with
menses
(past
2
months)
84
9.4
31
13.2
p=
.079
Hormonal
contraceptives
390
43.3
116
49.4
p
=
.098
(past
2
months)
Condom
use
(past
2
months)
none
276
36.2
88
44.0
p=
.087
£
5/10
times
172
22.6
31
15.5
6-9/10
times
126
16.5
35
17.5
10/10
times
188
24.7
46
23.0
Douched
(past
2
months)
433
48.2
69
29.4
p=
.000
Douching
frequency
Never
467
51.9
166
70.6
p=
.000
<
1
time/month
52
5.8
10
4.3
3
1
time/month
381
42.3'
59
25.1
Douching
recency
Never
467
51.9
166
70.6
p=
.000
0-7
days
150
16.7
13
5.5
8-21
days
124
13.8
28
11.9
322
days
158
17.6
28
11.9
JOURNAL
OF
THE
NATIONAL
MEDICAL
ASSOCIATION
VOL.
95,
NO.
3,
MARCH
2003
203
Identification
of
Bacterial
Vaginosis
At
baseline,
study
staff
taught
each
subject
how
to
self-collect
vaginal
specimens
using
a
Q-tip®-
like
cotton
swab.
Vaginal
swabs
were
smeared
onto
slides
by
study
staff
at
the
bedside
and
these
slides
were
air
dried
and
later
Gram-stained
at
a
centralized,
expert
microbiology
laboratory
under
the
direction
of
one
of
us
(SLH).
As
per
Nugent
et
al.,13
a
score
of
0-10
was
assigned
in
light
of
the
relative
proportions
of
large
Gram-positive
rods
(lactobacilli),
small
Gram-negative
or
Gram-vari-
able
rods
(Bacteroides
or
Gardnerella),
and
curved
Gram-variable
rods
(Mobiluncus).
A
score
of
0-3
was
interpreted
as
consistent
with
normal
vaginal
flora;
a
score
of
4-6,
corresponding
to
disturbed
flora,
was
designated
as
intermediate;
and
a
score
of
7-10
was
considered
to
be
bacterial
vaginosis.
Microbiologic
Methods
for
Evaluation
of
the
Vaginal
Flora
Two
swabs,
placed
in
an
anaerobic
transport
vial,
were
also
shipped
to
the
microbiology
lab-
oratory
for
characterization
of
the
bacteria
including
the
following:
Lactobacillus
species,
anaerobic
Gram
negative
rods,
Gardnerella
vaginalis,
and
Mycoplasma
hominis.
One
vagi-
nal
swab
was
used
to
inoculate
a
human
blood
bilayer
Tween
agar
for
detection
of
G.
vaginalis.
A
second
swab
was
used
to
inoculate a
Rogosa
agar
for
recovery
of
lactobacilli,
a
Brucella
agar
for
recovery
of
anaerobic
bacteria
and
broth
media
for
recovery
of
M.
hominis.
Lactobacilli
were
identified
to
the
genus
level
on
the
basis
of
Gram-stain
morphology
and
pro-
duction
of
lactic
acid.
All
lactobacilli
were
test-
ed
for
production
of
H202
using
a
qualitative
assay
on
a
tetramethylbenzidine
agar
plate.
After
two
days
of
incubation
at
370
C
in
the
anaerobic
glove
box,
the
agar
plates
were
exposed
to
ambi-
ent
air.
Hydrogen
peroxide
formed
and
horse
radish
peroxidase
in
the
agar
oxidized
the
tetramethylbenzidine,
which
caused
the
Lactobacillus
colonies
to
turn
blue.
Potential
Risk
Factors
In
a
standardized
20-minute
interview
con-
ducted
by
trained
research
staff
at
each
center,
women
were
asked
about
demographic
factors
including
age,
race,
highest
level
of
educational
attainment,
income,
marital
status,
and
gravidity.
They
reported
relevant
lifestyle
behaviors
such
as
tobacco
smoking,
number
of
sexual
partners
in
the
past
two
months,
acquisition
of
a
new
part-
ner
in
the
past
two
months,
various
types
of
con-
traceptive
use,
and
sex
with
menses.
Furthermore,
they
were
asked
to
recall
past
episodes
of
sexually
transmitted
infections
including
PID,
gonococcal
cervicitis,
chlamydial
cervicitis,
bacterial
vaginosis,
and
trichomonia-
sis.
Women
also
were
asked
about
their
douching
behavior
including
whether
they
had
practiced
any
douching
in
the
past
two
months,
their
fre-
quency
of
douching
in
the
past
two
months,
and
the
timing
of
the
most
recent
episode
of
douch-
ing.
For
purposes
of
analyses,
the
douching
fre-
quency
data
were
categorized
into
none,
<1/mo,
.1/mo,
and
the
douching
recency
data
were
cat-
egorized
into
seven
or
fewer
days
ago,
8-21
days
ago,
and
.22
days
ago.
Statistical
Analysis
The
differences
between
the
two
groups
were
analyzed
with
the
chi-square
test
for
categorical
comparisons.
First,
among
black
women
and
among
white
women
separately,
we
calculated
the
proportion
within
a
priori
risk
factor
cate-
gories.
Next,
within
each
racial
group,
we
evalu-
ated
the
relationship
between
risk
factors
and
BV.
Intermediate
vaginal
flora
was
combined
with
BV
into
a
single
outcome
because
the
two
categories
related
similarly
to
various
risk
fac-
tors
including
douching.
We
subsequently
com-
pared
BV
with
the
combination
of
intermediate
flora
plus
normal
flora
and
found
little
difference
in
the
main
results.
Odds
ratios,
with
correspon-
ding
95%
confidence
intervals
(CI)
were
calcu-
lated
as
the
primary
measure
of
effect
size
for
204
JOURNAL
OF
THE
NATIONAL
MEDICAL
ASSOCIATION
VOL.
95.
NO.
3,
MARCH
2003
Table
2.
UNADJUSTED
ODDS
RATIOS
(ORS)
FOR
BACTERIAL
VAGINOSIS/INTERMEDIATE
VAGINAL
FLORA
BY
RACE
Blacks
Whites
BV/lnt
Normal
OR(95%
Cl)
BVAnt
Normal
OR(95%
Cl)
Age
25+
19-24
13-18
Education
>HS
graduate
HS
graduate
<HS
graduate
189
382
41
282
197
133
Marital
status
Living
as
married
32
Never
married
552
Divorced/Widowed/Other
28
Income
3$20,000
92
$10,000-19,000
158
<$1
0,000
311
History
of
PID
No
509
Yes
102
History
of
Chiamydia
No
337
Yes
271
History
of
BV
No
Yes
347
254
History
of
Gonorrhea
No
442
Yes
165
History
of
Trichomoniasis
No
409
Yes
198
Gravidity
Never
205
Ever
407
Tobacco
smoking
Never
362
Current
207
Former
42
Number
of
partners
None
84
One
434
32
94
New
partners
No
Yes
Anal
sex
No
Yes
Sex
with
menses
No
Yes
431
97
575
37
1.0
0.9
(0.7-1.2)
0.8
(0.5-1.6)
1.0
0.8
(0.6-1.1)
0.7
(0.5-1.1)
1.0
1.0
(0.5-1.9)
0.8
(0.3-1.8)
1.0
0.7
(0.5-1.1)
0.6
(0.4-0.9)
93
165
17
146
79
50
14
245
16
59
74
114
239
33
161
113
175
96
216
59
219
54
108
167
181
70
24
50
196
29
198
27
266
9
536
74
Hormonal
contraception
No
371
Yes
241
Condom
use
None
£5/10
times
6-9/1
0
times
1
0/10
times
186
128
90
124
Douched
in
past
2
months
No
303
Yes
309
Douching
frequency
Never
303
<1
time/month
32
3
1
time/month
277
Douching
recency
Never
303
0-7
days
117
8-21
days
85
322
days
106
265
10
131
144
86
42
34
63
156
119
156
19
100
156
31
38
50
12
87
9
60
24
24
3
97
7
40
21
33
1.0
96
1.5
(1.0-2.2)
12
1.0
79
1.2
(0.9-1.5)
28
1.0
80
1.3
(1.0-1.8)
27
1.0
94
1.4
(1.0-1.9)
13
1.0
89
2.0
(1.4-2.8)
19
1.0
78
1.3
(1.0-1.7)
30
1.0
1.5
(1.1-2.0
0.9
(0.5-1.5
1.0
1.3
0.9-1.9
1.9
(1.1-3.3)
33
59
16
13
69
26
1.0
62
1.7
(1.0-2.6)
33
1.0
1.9
(0.9-4.0)
101
7
1.0
91
3.7
(1.9-7.2)
16
1.0
69
0.6
(0.4-0.8)
39
1.0
1.4
(0.9-2.2)
1.2
(0.8-2.0)
0.9
(0.6-1.4)
44
15
15
21
1.0
66
1.3
(1.0-1.8)
42
1.0
0.9
(0.5-1.6)
1.4
(1.1-1.9)
1.0
1.9(1.3-3.0)
1.2
(0.8-1.8)
1.1
(0.7-1.6)
66
6
36
66
8
20
14
16
100
8
98
16
10
3
117
4
77
12
26
117
6
102
21
100
24
117
6
118
6
103
21
46
53
25
22
72
30
65
37
112
12
1.0
0.9
(0.4-1.9)
0.7
(0.2-2.2)
1.0
0.4
(0.2-0.8)
0.3
(0.1-0.6)
1.0
1.2
(0.2-6.1)
1.8
(0.2-13.2)
1.0
0.3
(0.1-0.7)
0.4
(0.2-0.8)
1.0
2.4
(0.9-6.7)
1.0
1.7
(0.9-3.3)
1.0
1.4
(0.8-2.6)
1.0
2.7
(1.0-7.4)
1.0
4.2
(1.6-10.9)
1.0
1.9
(1.0-3.5)
1.0
1.6
(0.9-2.8
0.9
(0.4-1.9
1.0
1.6
(0.8-3.5)
1.5
(0.6-3.5)
1.0
0.9
(0.5-1.7)
1.0
0.7
(0.3-1.7)
1.0
1.4
(0.6-2.9)
1.0
0.4
(0.2-0.6)
1.0
1.1
(0.5-2.4)
0.7
(0.3-1.6)
0.8
(0.4-1.7)
1.0
2.5
(1.4-4.5)
1.0
2.3
(0.6-8.3)
2.6
(1.4-4.8)
1.0
2.4
(0.8-7.7)
110
14
48
76
43
14
20
25
99
25
99
4
21
99
5
8
12
3.8
(1.6-9.0)
1.8
(0.8-4.0)
BACTERIAL
VAGINOSIS
the relationship
between
risk
factors
and
BV/intermediate
vaginal
flora
within
racial
groups.
We
compared
odds
ratios
across
racial
groups
using
Woolf's
test
for
heterogeneity.
14
In
order
to
delineate
the
independent
rela-
tionships
between
risk
factors
and
the
dependent
variable,
BV/intermediate
vaginal
flora,
we
con-
ducted
separate
logistic
regression
models
with-
in
black
women
and
white
women.
Independent
variables
that
had
been
identified
on
the
basis
of
biologic
plausibility
and
preliminary
univariate
analyses
as
possible
confounders
were
included
in
these
models
with
the
exception
of
highly
collinear
variables.
We
also
substituted
educa-
tion
for
income,
as
the
later
was
missing
for
almost
10%
of
the
cohort
and
the
switch
of
edu-
cation
for
income
within
models
had
little
impact
on
the
significance
of
other
comparisons.
Simplification
of
the
models,
with
elimination
of
consistently
insignificant
covariates
(p<O.
10)
yielded
a
final
model
including:
clinical
site;
age
(continuous);
education
(high
school gradu-
ate
or
lesser
attainment
versus
post-high
school);
history
of
trichomoniasis
(yes/no);
gra-
vidity
(ever/never);
current
smoking
(yes/no);
sex
with
menses
(yes/no);
oral
contraceptive
or
other
hormonal
contraceptive
use
within
the
past
two
months
(yes/no);
and
douching
(never/<1/month/>1/month).
Finally,
we
com-
bined
all
(black
and
white)
women
into
one
model
and,
in
addition
to
all
other
terms,
includ-
ed
a
dichotomous
term
for
race.
This
allowed
us
to
determine
whether
race
remained
independ-
ently
associated
with
BV/intermediate
flora
after
adjustment
for
known
risk