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

net/publication/309695031

Association of Bacterial Vaginosis With Chlamydia and Gonorrhea Among

Women in the U.S. Army

Article  in  American Journal of Preventive Medicine · May 2017

DOI: 10.1016/j.amepre.2016.09.016

CITATIONS READS

19 713

6 authors, including:

Marcos Christian Bautista Eyako K Wurapa

Armed Forces Health Surveillance Division/DHA Walter Reed Army Institute of Research
127 PUBLICATIONS   3,425 CITATIONS    34 PUBLICATIONS   341 CITATIONS   

SEE PROFILE SEE PROFILE

Warren Sateren Sara M Morris

McGill University Lancaster University
49 PUBLICATIONS   1,311 CITATIONS    45 PUBLICATIONS   2,098 CITATIONS   

SEE PROFILE SEE PROFILE

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

psychosocial needs of cancer patients and their carers View project

CAMelot View project

All content following this page was uploaded by Jose Luis Sanchez Jr. on 25 August 2017.

The user has requested enhancement of the downloaded file.

 RESEARCH ARTICLE

Association of Bacterial Vaginosis With Chlamydia and

Gonorrhea Among Women in the U.S. Army
Christian T. Bautista, MSc, MPH,1 Eyako K. Wurapa, MD, MTM&H,2 Warren B. Sateren, MPH, MBA,2
Sara M. Morris, PhD,1 Bruce P. Hollingsworth, PhD,1 Jose L. Sanchez, MD, MPH3

Introduction: Bacterial vaginosis (BV) is a common vaginal condition in women of reproductive

age, which has been associated with Chlamydia trachomatis and Neisseria gonorrhoeae among
commercial sex workers and women attending sexually transmitted infection clinics. Pathogen-
specific associations between BV and other sexually transmitted infections among U.S. military
women have not been investigated.
Methods: A population-based, nested case-control study was conducted of all incident chlamydia
and gonorrhea cases reported to the Defense Medical Surveillance System during 2006
2012. Using
a density sampling approach, for each chlamydia or gonorrhea case, 10 age-matched (⫾1 year)
controls were randomly selected from those women who were never diagnosed with these infections.
Incidence rate ratios were estimated using conditional logistic regression. Statistical analysis was
carried out in December 2015.
Results: A total of 37,149 chlamydia cases and 4,987 gonorrhea cases were identified during the
study period. Antecedent BV was associated with an increased risk of subsequent chlamydia
(adjusted incidence rate ratio¼1.51; 95% CI¼1.47, 1.55) and gonorrhea (adjusted incidence rate
ratio¼2.42; 95% CI¼2.27, 2.57) infections. For every one additional episode of BV, the risk of
acquiring chlamydia and gonorrhea infections increased by 13% and 26%, respectively. A
monotonic dose
response relationship was also noted between antecedent BV and subsequent
chlamydia and gonorrhea infection. In addition, an effect modification on the additive scale was
found between BV and African-American race for gonorrhea, but not for chlamydia.

Conclusions: Among U.S. Army women, antecedent BV is associated with an increased risk of
subsequent chlamydia and gonorrhea infection.
Am J Prev Med 2017;52(5):632–639. & 2016 American Journal of Preventive Medicine. Published by Elsevier Inc.
This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).

INTRODUCTION New or multiple male sexual partners have been

associated with BV.4 Lower age at first intercourse, past

B
acterial vaginosis (BV) is a common vaginal
condition characterized by gray/white, thin, and
malodorous discharge.1 BV is associated with the
replacement of normal vaginal lactobacilli by an over-
growth of vaginal anaerobes or Gram-negative bacteria
(e.g., Gardnerella vaginalis), which causes an imbalance
in the vaginal microflora with a resulting discharge. The
BV prevalence varies substantially between countries
worldwide.2 In the U.S., BV is more frequently reported
among African-American and Hispanic women than
white women.3
pregnancy, commercial sex work, and vaginal douching
From the 1Faculty of Health and Medicine, Lancaster University, Lancaster,
United Kingdom; 2Walter Reed Army Institute of Research, Silver Spring,
Maryland; and 3Armed Forces Health Surveillance Branch, Public Health
Division, Defense Health Agency, Silver Spring, Maryland
Address correspondence to: Christian T. Bautista, MSc, MPH, Faculty of
Health and Medicine, Lancaster University, Bailrigg, Lancaster LA1 4YG,
Lancaster, United Kingdom. E-mail: cbautistat@gmail.com, m.bautista@
lancaster.ac.uk.
0749-3797/$36.00
http://dx.doi.org/10.1016/j.amepre.2016.09.016

632 Am J Prev Med 2017;52(5):632–639 & 2016 American Journal of Preventive Medicine. Published by Elsevier Inc. This is an
open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
 Bautista et al / Am J Prev Med 2017;52(5):632–639
5–8
are also linked to BV. Since the work of Leopold and 9 2012, were selected. For case findings, the ICD-9 was applied. For
Gardner and Dukes in the 1950s,10 the etiologic agent of
BV has not been established, and there is debate on
whether BV constitutes a sexually transmitted infection
(STI).11 However, it is also possible that BV is a sexually
enhanced disease.6
Data indicate that BV is a risk factor for the acquisition
of pathogens such as HIV, herpes simplex virus Type 2,
human papillomavirus, Chlamydia trachomatis, and
Neisseria gonorrhoeae.12–17 For example, one cohort
study conducted among non-pregnant women found
that those with BV had a 1.8- and 1.9-fold increased risk
for gonorrhea and chlamydia, respectively.18 A secondary
data analysis of the Project Protect clinical trial indicated
that women with a high BV severity had a 2.7-fold
increased risk of acquiring an STI.19
According to the Armed Forces Health Surveillance
Center, from 2000 through 2012, a total of 87,462
chlamydia and 11,403 gonorrhea cases were reported
among military women.20 Chlamydia and gonorrhea rates
are higher in military than civilian women. This difference
may be due to the high prevalence of risky sexual practices
among the predominantly younger military women. How-
ever, it might also be the result of a screening effect, as
women undergo routine annual screening for STIs.
The relationship between BV and STIs has not been
examined among military women. Thus, this study
sought to determine whether antecedent BV was asso-
ciated with subsequent chlamydia or gonorrhea infec-
tion, with special attention paid to Hispanic women, a
minority group not previously evaluated. This study also
analyzed the role of the number of episodes of BV on STI
risk, as a dose
response relationship has not previously
been studied. The identification of BV as risk factor for
the transmission of these infections has potential impli-
cations for prevention efforts in the U.S. military.
METHODS
Data Sample
Study data were obtained from the Defense Medical Surveillance
System (DMSS); details about this surveillance system have been
published elsewhere.21 Briefly, DMSS is the central repository of
medical surveillance data for the U.S. Armed Forces. This rela-
tional database contains information on personnel demographics,
reportable diseases, deployments, and medical data for all active-
duty personnel throughout their careers. DMSS data have pre-
viously been used to study HIV and herpes simplex virus Type 2
infections among U.S. military personnel.22,23
In this population-based, nested case-control study, all active
duty women in the U.S. Army with a first-time diagnosis of
chlamydia or gonorrhea, in either the first or second diagnostic
position of an outpatient or inpatient, physician-based, diagnostic
encounter reported between January 1, 2006 and December 31,
May 2017
633
chlamydia cases, the set ICD-9 codes were 099.41 or 099.5, and for
gonorrhea cases, the set ICD-9 codes were 098.0x, 098.1x, 098.4x,
or 098.8x. Using a density sampling approach,24 a random sample
of ten controls composed of women without a chlamydia or
gonorrhea diagnosis at the time of diagnosis for the case were
matched to each case on age (⫾1 year). Ten controls per case were
chosen to increase the statistical precision of the estimates.25
A BV diagnosis was based on the ICD-9-CM code 616.10
(vaginitis and vulvovaginitis, unspecified) in either the first or the
second diagnostic position of an outpatient or inpatient encounter
medical record from DMSS, prior to the diagnosis date of
chlamydia or gonorrhea. Because recurrent BV episodes are
frequent and there is no accepted definition of recurrence,26 an
episode of recurrent BV was defined as one where a subsequent
diagnosis took place Z30 days from the previously diagnosed BV
episode. This criterion was based on the recommendations given
by the Centers for Disease Control and Prevention BV Working
Group meta-analysis and synthesis report.3
Measures
Demographic variables, including race, education, marital status,
region of birth, rank, and years in service, were extracted from
DMSS. These were defined as race/ethnicity (white, African
American, Hispanic, and other), education (no high school, high
school, and some college or higher), marital status (single, married,
and other), rank (lower enlisted [E1
E4], higher enlisted E5
E9],
and officers), and years of service (r1, 2
3, and Z4). Geographic
region or residence (birth) was classified into four regions (West,
Midwest, Northeast, and South).
Statistical Analysis
The prevalence of BV among cases and controls was compared
using the Cochran
Mantel
Haenszel test. To analyze individual
matched case-control data, conditional logistic regression was used
to estimate the incidence rate ratios with 95% CIs to determine the
association between BV and chlamydia and gonorrhea infection.
Separate regression models were performed for each variable
category. In addition to analyzing BV as a binary variable, BV
was treated as a categorical variable to conduct a dose
response
relationship analysis. An effect modification analysis on the
additive scale was conducted to determine whether the effect of
BV on chlamydia or gonorrhea infection differed by race/ethnicity.
For analysis, the synergy and the relative excess risk due to
interaction [RERI] measures were calculated. Synergy equal to one
and RERI equal to zero indicate there is no departure from
additivity.25
All p-values were two-sided and were considered statistically
significant if po0.05. Analyses were carried out in December 2015
using Stata, version 14.0.
The study was approved by scientific review and IRBs at
Lancaster University and the Walter Reed Army Institute of
Research.
RESULTS
Chlamydia
Between 2006 and 2012, a total of 37,149 chlamydia case
patients were identified among U.S. Army women. Of
634 Bautista et al / Am J Prev Med 2017;52(5):632–639
Table 1. BV Prevalence Among Chlamydia and Gonorrhea Cases and Matched Controls Among U.S. Army Females

Chlamydia Gonorrhea

Cases Controls Cases Controls

Feature BV % n/N BV % n/N BV % n/N BV % n/N

All participants 21.4 7,942/37,149 15.6 57,793/371,490 34.7 1,730/4,987 19.1 9,523/49,870
Race/ethnicity
White 16.5 2,440/14,827 12.2 23,623/193,291 26.1 297/1,136 14.2 3,608/25,096
African- 29.1 3,715/12,753 26.2 20,326/77,712 39.4 1,119/2,838 31.8 3,681/11,571
American
Hispanic 20.0 996/4,990 14.9 7,732/51,949 32.0 168/525 18.5 1,252/6,777
Othera 15.7 623/3,961 11.4 4,700/41,236 27.1 106/391 13.6 740/5,426
Education level
No high school 17.4 24/138 13.4 167/1,244 15.0 3/20 19.7 33/167
High school 20.5 6,758/32,929 14.9 46,727/312,692 34.1 1,521/4,456 18.4 7,602/41,330
College or 29.7 1,060/3,564 20.3 10,160/50,058 41.7 183/439 24.1 1,766/7,325
higher
Marital status
Single 16.6 4,211/25,371 11.2 26,653/238,865 28.2 919/3,257 13.8 4,181/30,369
Married 29.3 2,748/9,373 22.3 26,398/118,436 44.2 605/1,370 25.6 4,376/17,107
Otherb 40.9 976/2,387 33.7 4,712/13,993 57.1 201/352 40.6 960/2,367
Region of birthc
West 17.6 1,233/7,003 12.6 10,694/84,997 29.1 196/673 15.2 1,680/11,031
Midwest 19.3 1,021/5,279 13.8 7,995/57,716 35.3 211/598 16.6 1,246/7,519
Northeast 22.5 810/3,605 15.6 6,427/41,169 36.8 164/445 19.4 1,040/5,355
South 23.7 4,187/17,690 17.8 26,608/149,207 35.5 1,033/2,906 22.0 4,468/20,343
Military rank
Junior enlisted 17.7 5,586/31,579 12.1 35,149/289,502 30.2 1,239/4,107 14.6 5,420/37,029
Senior enlisted 46.0 2,091/4,544 35.3 18,884/53,429 59.4 458/771 38.8 3,470/8,931
Officers 25.8 265/1,026 13.2 3,760/28,559 30.3 33/109 16.2 633/3,910
Years of military service
r1 9.8 1,784/18,148 5.4 8,383/154,241 16.3 339/2,072 6.6 1,191/18,126
2
3 23.7 2,364/9,954 15.4 17,562/114,279 37.8 508/1,342 17.2 2,611/15,203
Z4 41.9 3,794/9,047 30.9 31,848/102,970 56.2 883/1,572 34.6 5,721/16,541
a
Other: American Indian/Alaskan Native, Asian/Pacific Islander, and others.
b
Other: Divorced, widowed, and others.
c
West (MT, WY, CT, NM, ID, UT, AZ, NV, WA, OR, CA, AK, and HI); Midwest (IL, IN, IA, KS, MI, MN, MO, OH, NE, ND, SD, and WI); Northeast (ME, NH, VT,
MA, RI, CT, NY, NJ, and PA); South (FL, GA, NC, SC, VA, WV, MD, DC, DE, AL, KY, MS, TN, AR, LA, OK, and TX).
BV, bacterial vaginosis.

these, 89% had high school education, 85% were in the
lower enlisted rank, 80% were aged o25 years, 68% were
single, 49% had at least 1 year of military service, 48% had
home of record in the South, and 34% were African
American. Among the 371,490 controls, 84% had high
school education, 78% were of lower enlisted rank, 64%
were single, 41% had at least 1 year of military service, 40%
were from the South, and 21% were African American.
The prevalence of BV was higher among chlamydia
cases than their controls (21.4% vs 15.6%, po0.001)
(Table 1). For both chlamydia cases and controls, the
highest BV prevalences occurred among women with
senior enlisted rank, those with 43 years of service, those
with “other” marital status, women with college or higher
education, married personnel, and among African
Americans.
After adjusting for covariates, BV was significantly
associated with chlamydia infection (incidence rate
ratio¼1.51; po0.001) (Table 2). These analyses also
showed that the association between BV and chlamydia
was higher in women with junior enlisted rank, women
who had r2 years of military service, women with single
marital status, and who were from the Northeast.
Interestingly, the association between BV and chlamydia
was higher among white women compared with African-
American and Hispanic women.

www.ajpmonline.org
 Bautista et al / Am J Prev Med 2017;52(5):632–639 635
Table 2. Crude and Adjusted Associations Between BV and Chlamydia and Gonorrhea Among U.S. Army Females

Chlamydia Gonorrhea

Feature IRR (95% CI) Adjusteda IRR (95% CI) IRR (95% CI) Adjusteda IRR (95% CI)
All participants 1.52 (1.47, 1.56) 1.51 (1.47, 1.55) 2.42 (2.26, 2.58) 2.42 (2.27, 2.57)
Race/ethnicity
White 1.46 (1.39, 1.54) 1.43 (1.36, 1.50) 2.28 (1.94, 2.68) 2.31 (1.96, 2.71)
African American 1.22 (1.16, 1.29) 1.24 (1.17, 1.31) 1.56 (1.41, 1.73) 1.58 (1.42, 1.75)
Hispanic 1.33 (1.20, 1.49) 1.36 (1.22, 1.52) 2.08 (1.53, 2.82) 2.12 (1.55, 2.90)
Otherb 1.33 (1.15, 1.53) 1.36 (1.18, 1.57) 2.65 (1.76, 3.99) 2.68 (1.78, 4.03)
Education level
No high school Undefinedc Undefinedc Undefinedc Undefinedc
High school 1.44 (1.40, 1.49) 1.50 (1.46, 1.55) 2.32 (2.16, 2.49) 2.44 (2.27, 2.62)
College or higher 1.68 (1.53, 1.84) 1.44 (1.31, 1.59) 2.20 (1.72, 2.81) 1.90 (1.47, 2.46)
Marital status
Single 1.63 (1.57, 1.69) 1.59 (1.53, 1.65) 2.71 (2.47, 2.97) 2.66 (2.42, 2.91)
Married 1.46 (1.38, 1.54) 1.44 (1.36, 1.51) 2.24 (1.97, 2.54) 2.23 (1.96, 2.54)
Otherd 1.38 (1.18, 1.61) 1.37 (1.17, 1.60) 1.68 (1.18, 2.41) 1.66 (1.15, 2.39)
Region of birthe
West 1.43 (1.32, 1.56) 1.42 (1.31, 1.55) 2.55 (1.99, 3.26) 2.59 (2.02, 3.32)
Midwest 1.51 (1.36, 1.68) 1.49 (1.34, 1.66) 2.90 (2.16, 3.90) 2.85 (2.10, 3.87)
Northeast 1.60 (1.40, 1.83) 1.52 (1.33, 1.75) 2.48 (1.75, 3.51) 2.41 (1.69, 3.43)
South 1.47 (1.41, 1.54) 1.48 (1.42, 1.54) 2.12 (1.93, 2.33) 2.14 (1.95, 2.36)
Military rank
Junior enlisted 2.20 (1.78, 2.71) 2.15 (1.74, 2.66) 2.37 (1.28, 4.37) 2.40 (1.29, 4.46)
Senior enlisted 1.47 (1.43, 1.52) 1.50 (1.45, 1.55) 2.46 (2.28, 2.66) 2.53 (2.34, 2.74)
Officers 1.56 (1.45, 1.67) 1.54 (1.43, 1.65) 2.22 (1.88, 2.63) 2.19 (1.85, 2.60)
Years of military service
r1 1.81 (1.71, 1.92) 1.82 (1.72, 1.93) 2.74 (2.36, 3.18) 2.74 (2.35, 3.19)
2
3 1.71 (1.61, 1.81) 1.72 (1.63, 1.83) 2.95 (2.56, 3.38) 2.95 (2.56, 3.40)
Z4 1.63 (1.56, 1.71) 1.61 (1.53, 1.69) 2.42 (2.16, 2.70) 2.38 (2.13, 2.68)
a
Adjusted IRR for race/ethnicity, marital status, and military rank.
b
Other: American Indian/Alaskan Native, Asian/Pacific Islander, and others.
c
Undefined IRR, zero or infinity.
d
Other: Divorced, widowed, and others.
e
West (MT, WY, CT, NM, ID, UT, AZ, NV, WA, OR, CA, AK, and HI); Midwest (IL, IN, IA, KS, MI, MN, MO, OH, NE, ND, SD, and WI); Northeast (ME, NH, VT,
MA, RI, CT, NY, NJ, and PA); South (FL, GA, NC, SC, VA, WV, MD, DC, DE, AL, KY, MS, TN, AR, LA, OK, and TX).
BV, bacterial vaginosis; IRR, incidence rate ratio.

When BV was treated as a continuous variable, for
every additional one episode of BV, the risk of chlamydia
infection increased by 13% (95% CI¼12%, 14%) com-
pared with women without BV. A similar estimate was
found after controlling for race/ethnicity, marital status,
and military rank (data not shown). When BV was
treated as a categorical predictor, a dose
response
relationship with chlamydia infection was found
(Figure 1). Women with one episode of BV had 1.45
(95% CI¼1.40, 1.50) times more risk of acquiring
chlamydia infection compared with women without
BV. Additionally, for women with two, three, four, five,
and six or more episodes of BV, the risk of acquiring
chlamydia compared with women without BV was 1.55
(95% CI¼1.47, 1.64), 1.74 (95% CI¼1.61, 1.88), 1.46
(95% CI¼1.29, 1.65), 1.95 (95% CI¼1.68, 2.26), and 2.07
(95% CI¼1.83, 2.33), times higher, respectively.
The analysis did not detect an additive effect mod-
ification between BV and chlamydia by race categorized
into two groups (African-American race and other race/
ethnic groups [white/Hispanic/other]; synergy¼1.02,
95% CI¼0.88, 1.90; RERI¼0.04, 95% CI¼
0.24, 0.36).
Gonorrhea
During the study period, a total of 4,987 gonorrhea case
patients were identified (Table 1). Of these, 89% had high
school education, 82% were lower enlisted rank, 76%
were aged o25 years, 65% were single, 58% were from
the South, 57% were African American, and 41% had at
least 1 year of military service. Among the 49,870

May 2017
636 Bautista et al / Am J Prev Med 2017;52(5):632–639
Figure 1. Associations for chlamydia infection by the
number of episodes of bacterial vaginosis among U.S. Army
females.
controls, 83% had high school education, 74% were lower
enlisted rank, 61% were single, 41% were from the South,
36% had at least 1 year of military service, and 23% were
African American.
The prevalence of BV was higher among gonorrhea
cases than their controls (34.7% vs 19.1%, po0.001)
(Table 1). Similar to the chlamydia results, the highest
BV prevalences among gonorrhea cases and controls
occurred among senior enlisted personnel, among
women with Z4 years of service, African-American
women, those with other marital status, and among
women with college or higher education.
In a regression analysis, BV was significantly associated
with gonorrhea (incidence rate ratio¼2.42; po0.001).
After controlling for race, marital status, and military
rank, the association between BV and gonorrhea was
higher in women with the following characteristics: those
with o4 years of military service, those from the Midwest,
single women, and those of other race/ethnic groups.
Analyses also revealed that for every additional epi-
sode of BV, the risk of gonorrhea infection increases 26%
(95% CI¼23%, 28%, po0.001) compared with women
without BV. This estimate was similar after controlling
for race/ethnicity, marital status, and military rank (data
not shown). When BV was analyzed as a categorical
predictor, the results showed a linear association between
the number of episodes of BV and acquisition of
gonorrhea infection (Figure 2). According to this
dose
response analysis, women with one episode of BV
had 1.98 (95% CI¼1.82, 2.15) times more risk of acquiring
gonorrhea compared with women without BV. Addi-
tionally, compared with women without BV, the risk of
acquiring gonorrhea among women with two, three, four,
five, and six or more episodes was 2.71 (95% CI¼2.41,
3.04), 3.22 (95% CI¼2.74, 3.79), 3.62 (95% CI¼2.87,
Figure 2. Associations for gonorrhea infection by the num-
ber of episodes of bacterial vaginosis among U.S. Army
females.
4.55), 4.12 (95% CI¼3.14, 5.40), and 4.62 (95% CI¼3.73,
5.73), respectively.
Analyses also found that using an additive scale, race/
ethnicity was a positive effect modifier for the relation-
ship between BV and gonorrhea (synergy¼1.29, 95%
CI¼1.05, 1.58; RERI¼1.80, 95% CI¼0.25, 3.34). This
synergistic relationship indicates that the joint exposure
effect of BV and African-American race on gonorrhea
was larger than the sum of the effect of African-American
race on gonorrhea risk and of BV on gonorrhea risk.
DISCUSSION
Antecedent BV was associated with subsequent chlamy-
dia and gonorrhea infection among women in the U.S.
Army. This represents the first large study to provide
objective evidence of the role of the most common cause
of vaginal discharge, BV, on these two STIs. This
association was previously reported only among high-
risk women in civilian populations.15–18
Although it is difficult to estimate the exact prevalence of
BV, as many women are asymptomatic, limited data on this
vaginal condition indicates that among women entering mili-
tary service in the U.S. Marine Corps during 1999
2000, a
high prevalence was found among African-American (32%)
and Hispanic women (30%).4 This finding is in agreement
with that observed in this study, in which African-American
and Hispanic women had the highest BV prevalences.
The prevalence of BV was two and three times higher
among African-American women with chlamydia or
gonorrhea compared with white women. This disparity,
which has been reported in civilian studies,3,16 cannot be
explained by differences in sexual activity or sociodemo-
graphic characteristics.27 It is possible that it is associated
www.ajpmonline.org
 Bautista et al / Am J Prev Med 2017;52(5):632–639
with lifestyle factors that are more common among
African-American women, such as vaginal douching.
Vaginal douching, which causes a disequilibrium in the
vaginal microflora, induces inflammation through either
physical or chemical irritation, and therefore predisposes
women to BV.28 On the other hand, it is hypothesized
that differences in the composition of bacterial vaginal
flora between whites and African Americans might
explain this disparity,29 but there is no direct evidence
of such a complex relationship.
This study found that women with BV were 1.5 and
2.4 times more likely to have a subsequent chlamydia or
gonorrhea diagnosis when compared with their aged-
matched controls, respectively. These estimates were
smaller for chlamydia, and higher for gonorrhea, when
compared with estimates from a large cohort study
among 3,620 non-pregnant women aged 15
44 years
(hazard ratio for chlamydia¼1.9; hazard ratio for gon-
orrhea¼1.8).18 These findings are consistent with pre-
vious research suggesting that BV facilitates the
acquisition of these STIs.
This study also documents that among Hispanic
women, the relationship of BV with chlamydia or
gonorrhea infection was higher when compared with
African-American women. Data on the dynamics of STI
transmission, including gonorrhea, among Hispanic
women is limited in the military literature, despite the
fact that this minority group has increased its represen-
tation in the U.S. military from 4% in 1981 to 13% in
2013.30 Further studies are warranted in the military to
determine what individual, social, or biological factors
contribute to the risk of STIs among Hispanic women.
Analyses revealed that among single and enlisted
women, the association of BV with chlamydia or gonor-
rhea was stronger compared with those who were
married or officers. Studies of active duty service women,
who are mostly single and young, report that they engage
in high-risk sexual behaviors (e.g., multiple sexual
partners, low condom use, and sex while under the
influence of drugs or alcohol).20,31 Additionally, among
young female recruits (aged 17
21 years), BV has been
associated with having more than one sexual partner in
the past 3 months.4 Considering these findings, it is valid
to assume that high-risk sexual behaviors might explain
the difference between single and married women. On
the other hand, in the U.S. Army, enlisted personnel are
predominantly young and unmarried. In this enlisted
population, 71% of women were young (aged o25 years)
and unmarried. Consequently, this high percentage may
explain, in part, the high association of BV with
chlamydia and gonorrhea among the enlisted ranks.
An important finding from this study was that for
every one additional episode of BV, the risk of acquiring
May 2017
637
chlamydia and gonorrhea infection increased by 13% and
26%, respectively. This increased risk was independent of
other covariates. To the best of the authors’ knowledge,
there is no evidence in the literature on the risk of STIs by
the number of episodes of BV. Although the basis for this
increased risk is not known, this study presents evidence
for the cumulative effect of BV on the risk for these
infections.
This study also found a dose
response relationship
between antecedent BV and subsequent chlamydia and
gonorrhea diagnosis. The shape of this relationship was
monotonic, which indicates that the risk of these infec-
tions increases with the number of episodes of BV. For
simplicity, the shape of the trend was analyzed descrip-
tively, and not using advanced statistical models, such as
splines.32 This dose
response relationship supports the
role of BV in the risk of STI acquisition. In addition to the
observed strength of the association, an important com-
ponent for disease causality, the presence of a dose

response relationship has public health implications in
terms of disease prevention because it provides clues to the
biology underlying bacterial
STI relationship.
Effect modification analyses found a synergistic effect
between BV and African-American race for gonorrhea,
but not for chlamydia. Although the mechanisms of this
effect remain unclear, biological factors, in addition to
individual behavioral and social factors, may explain racial
disparities in STI acquisition.27–29 However, the nature of
STIs among African Americans is complex, multifactorial
in nature, and requires further individual and social
network
level behavioral research. This is the first study
that suggests that African-American race modifies the
association between BV and gonorrhea infection.
Taken together, this study provides additional evidence
of a probable causal link between BV and chlamydia and
gonorrhea infection. These conclusions are supported by
the strength of the associations, the temporal sequence of
the associations, the consistency of findings with previous
studies, and a dose
response relationship.
Limitations
This study has some limitations. It is likely that the
reported number of women with BV was affected by the
limitations of use of ICD-9 codes as indicators of BV.
Therefore, it is probable that some women with BV were
misclassified or had other conditions that also cause an
inflammation or infection of the vagina, such as candi-
diasis or noninfectious vaginitis. It was not possible to
analyze behavioral risk data, given that this information
is not collected in DMSS. Further research should study
the influence of behavioral factors (e.g., sexual practices,
douching, use of condoms, and frequency of vaginal sex)
638 Bautista et al / Am J Prev Med 2017;52(5):632–639
on the association of BV with chlamydia and gonorrhea
among military women.
CONCLUSIONS
Among women in the U.S. Army, antecedent BV is
associated with an increased risk of subsequent chlamy-
dia and gonorrhea infection. This finding is consistent
with those observed amongst civilian populations where
BV has been reported to represent a strong risk factor for
both bacterial STIs. The dose
response relationship
between the number of episodes of BV and chlamydia
or gonorrhea infection highlights the importance of BV
as a clinical risk factor, which can be considered
“hypothesis generating” for further research in the field
of STIs. The study also provides important evidence-
based public health findings for U.S. military authorities
to develop prevention strategies such as sexual risk
reduction for female service members with BV in order
to reduce the acquisition of STIs. These results
strengthen the literature that BV constitutes a predis-
posing condition that increases the risk of both chlamy-
dia and gonorrhea infections.
ACKNOWLEDGMENTS
The authors would like to thank Dr. Angelia Cost, a senior
managing epidemiologist at the Armed Forces Health Surveil-
lance Branch, for Defense Medical Surveillance System data-
base extraction and epidemiologic support for this study, and to
Mr. Sebastian-Santiago for technical assistance.
This study was funded by the U.S. Armed Forces Health
Surveillance Branch and its Global Emerging Infectious Surveil-
lance section.
The views expressed herein are those of the authors and do
not reflect the official policy or position of the Department of the
Army, Department of Defense, the U.S. Government, or any
listed organization. Some authors are employees of the U.S.
government. This work was prepared as part of their official
duties and, as such, there is no copyright to be transferred. This
report was approved for publication by the U.S. Army Public
Health Command.
Preliminary results from this study were presented at the 29th
European Conference on Sexually Transmitted Infections,
poster #168, Barcelona, Spain, September 24
26, 2015.
No financial disclosures were reported by the authors of
this paper.
REFERENCES
1. Koumans EH, Sternberg M, Bruce C, et al. The prevalence of bacterial
vaginosis in the United States, 2001-2004; associations with symptoms,
sexual behaviors, and reproductive health. Sex Transm Dis. 2007;34
(11):864–869. http://dx.doi.org/10.1097/OLQ.0b013e318074e565.
2. Kenyon C, Colebunders R, Crucitti T. The global epidemiology of
bacterial vaginosis: a systematic review. Am J Obstet Gynecol. 2013;209
(6):505–523. http://dx.doi.org/10.1016/j.ajog.2013.05.006.
3. Koumans EH, Markowitz LE, Hogan V, CDC BV Working Group.
Indications for therapy and treatment recommendations for bacterial
vaginosis in nonpregnant and pregnant women: a synthesis of data.
Clin Infect Dis. 2002;35(suppl 2):S152–S172.
4. Fethers KA, Fairley CK, Hocking JS, Gurrin LC, Bradshaw CS. Sexual
risk factors and bacterial vaginosis: a systematic review and meta-
analysis. Clin Infect Dis. 2008;47(11):1426–1435. http://dx.doi.org/
10.1086/592974.
5. Yen S, Shafer MA, Moncada J, et al. Bacterial vaginosis in sexually
experienced and non-sexually experienced young women entering the
military. Obstet Gynecol. 2003;102(5 Pt 1):927–933. http://dx.doi.org/
10.1016/S0029-7844(03)00858-5.
6. Smart S, Singal A, Mindel A. Social and sexual risk factors for bacterial
vaginosis. Sex Transm Infect. 2004;80(1):58–62. http://dx.doi.org/
10.1136/sti.2003.004978.
7. Verstraelen H, Verhelst R, Vaneechoutte M, Temmerman M. The
epidemiology of bacterial vaginosis in relation to sexual behaviour.
BMC Infect Dis. 2010;10:81. http://dx.doi.org/10.1186/1471-2334-
10-81.
8. Fethers KA, Fairley CK, Morton A, et al. Early sexual experiences and
risk factors for bacterial vaginosis. J Infect Dis. 2009;200(11):1662–
1670. http://dx.doi.org/10.1086/648092.
9. Leopold S. Heretofore undescribed organism isolated from the genito-
urinary system. U.S. Armed Forces Med. 1953;4(2):263–266.
10. Gardner HL, Dukes CD. Haemophilus vaginalis vaginitis: a newly
defined specific infection previously classified non-specific vaginitis.
Am J Obstet Gynecol. 1955;69(5):962–976. http://dx.doi.org/10.1016/
0002-9378(55)90095-8.
11. Turovskiy Y, Noll KS, Chikindas ML. The aetiology of bacterial
vaginosis. J Appl Microbiol. 2011;110(5):1105–1128. http://dx.doi.org/
10.1111/j.1365-2672.2011.04977.x.
12. Leppäluoto PA. Autopsy of bacterial vaginosis: a physiological entity
rather than a contagious disease. Acta Obstet Gynecol Scand. 2008;87
(5):578–579. http://dx.doi.org/10.1080/00016340802031049.
13. Mirmonsef P, Krass L, Landay A, Spear GT. The role of bacterial
vaginosis and trichomonas in HIV transmission across the female
genital tract. Curr HIV Res. 2012;10(3):202–210. http://dx.doi.org/
10.2174/157016212800618165.
14. Nagot N, Quedraogo A, Defer MC, et al. Association between bacterial
vaginosis and Herpes simplex virus type-2 infection: implications for
HIV acquisition studies. Sex Transm Infect. 2007;83(5):365–368. http:
//dx.doi.org/10.1136/sti.2007.024794.
15. Brotman RM, Shardell MD, Gajer P, et al. Interplay between the
temporal dynamics of the vaginal microbiota and human papilloma-
virus detection. J Infect Dis. 2014;210(11):1723–1733. http://dx.doi.org/
10.1093/infdis/jiu330.
16. Wiesenfeld HC, Hillier SL, Krohn MA, Landers DV, Sweet RL.
Bacterial vaginosis is a strong predictor of Neisseria gonorrhoeae and
Chlamydia trachomatis infection. Clin Infect Dis. 2003;36(5):663–668.
http://dx.doi.org/10.1086/367658.
17. Gallo MF, Macaluso M, Warner L, et al. Bacterial vaginosis, gonorrhea,
and chlamydial infection among women attending a sexually trans-
mitted disease clinic: a longitudinal analysis of possible causal links.
Ann Epidemiol. 2012;22(3):213–220. http://dx.doi.org/10.1016/j.
annepidem.2011.11.005.
18. Brotman RM, Klebanoff MA, Nansel TR, et al. Bacterial vaginosis
assessed by gram stain and diminished colonization resistance to
incident gonococcal, chlamydial, and trichomonal genital infection. J
Infect Dis. 2010;202(12):1907–1915. http://dx.doi.org/10.1086/657320.
19. Allsworth JE, Peipert JF. Severity of bacterial vaginosis and the risk of
sexually transmitted infection. Am J Obstet Gynecol. 2011;205(2):113.
http://dx.doi.org/10.1016/j.ajog.2011.02.060.
20. Armed Forces Health Surveillance Center. Sexually transmitted
infections, active component, U.S. Armed Forces, 2000
2012. MSMR.
2013;20(2):5–10.
www.ajpmonline.org
 Bautista et al / Am J Prev Med 2017;52(5):632–639
21. Rubertone MV, Brundage JF. The Defense Medical Surveillance
System and the Department of Defense serum repository: glimpses
of the future of public health surveillance. Am J Public Health. 2002;92
(2):1900–1904. http://dx.doi.org/10.2105/AJPH.92.12.1900.
22. Bautista CT, Sateren WB, Sanchez JL, et al. HIV incidence trends
among white and African-American active duty United States Army
personnel (1986
2003). J Acquir Immune Defic Syndr. 2006;43(3):
351–355. http://dx.doi.org/10.1097/01.qai.0000243051.35204.d0.
23. Bautista CT, Singer DE, O’Connell RJ, et al. Herpes simplex virus type
2 and HIV infection among U.S. military personnel: implications for
health prevention programmes. Int J STD AIDS. 2009;20(9):634–637.
http://dx.doi.org/10.1258/ijsa.2008.008413.
24. Richardson D. An incidence density sampling program for nested case-
control analyses. Occup Environ Med. 2004;61:e59. http://dx.doi.org/
10.1136/oem.2004.014472.
25. Rothman KJ, Greenland S, Lash TL, eds. Modern Epidemiology.
Philadelphia, PA: Lippincott Williams & Wilkins, 2008.
26. Eschenbach DA. Bacterial vaginosis: resistance, recurrence, and/or
reinfection? Clin Infect Dis. 2007;44(2):220–221. http://dx.doi.org/
10.1086/509584.
May 2017
View publication stats
639
27. Royce RA, Jackson TP, Thorp JM Jr, et al. Race/ethnicity, vaginal flora
patterns, and pH during pregnancy. Sex Transm Dis. 1999;26(2):96–
102. http://dx.doi.org/10.1097/00007435-199902000-00007.
28. Brotman RM, Klebanoff MA, Nansel TR, et al. A longitudinal study of
vaginal douching and bacterial vaginosis: a marginal structural
modeling analysis. Am J Epidemiol. 2008;168(2):188–196. http://dx.
doi.org/10.1093/aje/kwn103.
29. Hickey RJ, Zhou X, Pierson JD, Ravel J, Forney LJ. Understanding vaginal
microbiome complexity from an ecological perspective. Transl Res.
2012;160(4):267–282. http://dx.doi.org/10.1016/j.trsl.2012.02.008.
30. Holder K. Post 9/11 women veterans. Proceedings of the Annual
Meeting of the Population Association of America; Apr 15
17;2010.
Dallas, TX: Housing and Household Economic Statistics Division,
U.S. Census Bureau; 2010.
31. Stahlman S, Javanbakht M, Cochran S, et al. Self-reported STIs and
sexual risk behaviors in the U.S. military: how gender influences risk.
Sex Transm Dis. 2014;41(6):359–364. http://dx.doi.org/10.1097/
OLQ.0000000000000133.
32. Steenland K, Deddens JA. A practical guide to dose-response
analyses and risk assessment in occupational epidemiology. Epidemio-
logy. 2004;15(1):63–70. http://dx.doi.org/10.1097/01.ede.0000100287.
45004.e7.