Académique Documents
Professionnel Documents
Culture Documents
VOL. 109, NO. 1, JANUARY 2007 Allsworth and Peipert Prevalence of Bacterial Vaginosis 115
Table 1. Prevalence of Bacterial Vaginosis and 95% Confidence Interval Among Women 14 – 49 Years of
Age by Sociodemographic Characteristics and Reproductive History
Prevalence (95% Confidence Interval)
116 Allsworth and Peipert Prevalence of Bacterial Vaginosis OBSTETRICS & GYNECOLOGY
United States was high—almost one woman in three rial vaginosis than those who did not (45% versus
was positive for bacterial vaginosis (29.2%). Although 24%; P⬍.001).
younger women, those between the ages of 14 and 19 We examined the association between sexual
years, had a somewhat lower prevalence (23.3%) of history and bacterial vaginosis prevalence among
bacterial vaginosis, among the 20 years and older women between the ages of 20 and 49 years and
group the prevalence was between 28% and 31%. The found that a history of sexual intercourse and number
prevalence of bacterial vaginosis varied significantly of male sexual partners were associated with bacterial
with race or ethnicity, education, and poverty/income vaginosis. Women who reported no history of sexual
ratio. Bacterial vaginosis was more common among intercourse had rates of bacterial vaginosis that were
black, non-Hispanic (51.6%) and Mexican-American half that of women who reported a history of inter-
(32.1%) women than among white, non-Hispanic course (15% versus 30%). The prevalence of bacterial
women (23.2%). Women with more than a high vaginosis was highest among those who reported the
school education were less likely to be positive for youngest ages of first intercourse: 38% among those
bacterial vaginosis than those with a high school who reported first intercourse between the ages of 14
education or less (26% versus 33–34%). Similarly, the and 19 compared with 30% for those aged 15–19
prevalence of bacterial vaginosis was lower among years and 24% for those whose first intercourse oc-
those living well above the federal poverty level (24%) curred at age 20 or older. Number of male sexual
compared with those living at or near (34%) or below partners during one’s lifetime and in the past year was
(37%) the federal poverty level. also associated with prevalence of bacterial vaginosis.
Selected reproductive history characteristics were Prevalence was lowest among women with the fewest
also associated with the prevalence of bacterial vagi- male sexual partners in their lifetimes and among
nosis. Although there was no noticeable trend in the those with a single male sexual partner in the past
relationship to bacterial vaginosis and age at men- year.
arche, women who reported douching in the past 6 Results from crude and adjusted logistic regres-
months had significantly higher prevalence of bacte- sion analyses are presented in Table 2. Although the
Table 2. Crude and Adjusted Odds Ratios and 95% Confidence Intervals for Selected Demographic
Characteristics Among Women 14 – 49 Years of Age
Model 2:
Model 1: Adjusted for
Adjusts for Sociodemographics
Crude Sociodemographics and Douching
Age (y)
14–19 0.67 (0.52–0.87) 0.56 (0.40–0.80) 0.65 (0.46–0.93)
20–29 — — —
30–39 0.87 (0.71–1.05) 0.95 (0.78–1.15) 0.91 (0.76–1.10)
40–49 1.01 (0.74–1.37) 1.17 (0.88–1.56) 1.12 (0.85–1.49)
Race or ethnicity
White, non-Hispanic — — —
Black, non-Hispanic 3.52 (2.97–4.19) 3.13 (2.58–3.80) 2.66 (2.18–3.25)
Mexican American 1.57 (1.22–2.01) 1.29 (0.99–1.69) 1.33 (1.03–1.72)
Other race/Hispanic 1.89 (1.22–2.91) 1.76 (1.13–2.75) 1.78 (1.11–2.85)
Education
Less than high school 1.42 (1.17–1.73) 1.47 (1.13–1.92) 1.37 (1.05–1.77)
High school/GED 1.48 (1.17–1.88) 1.38 (1.08–1.75) 1.29 (0.99–1.66)
More than high school — — —
Household income in relation to federal poverty below
Below (PIR 0 to 1.0) 1.84 (1.46–2.31) 1.43 (1.17–1.74) 1.32 (1.08–1.63)
At (PIR more than 1.0 to 2.0) 1.63 (1.25–2.13) 1.34 (1.02–1.77) 1.27 (0.95–1.69)
Above (PIR 2.0 or more) — — —
Douched in past 6 months
No — — —
Yes 2.57 (2.19–3.03) — 1.93 (1.54–2.40)
Missing 2.33 (1.72–3.17) — 1.90 (1.36–2.65)
GED, general equivalency diploma; PIR, poverty/income ratio.
VOL. 109, NO. 1, JANUARY 2007 Allsworth and Peipert Prevalence of Bacterial Vaginosis 117
odds of bacterial vaginosis were lower among women
between the ages of 14 and 19 years compared with
20 –29 year olds, there were no significant differences
between 20 –29 year olds and those at ages 30 –39
years or 40 – 49 years. Differences observed in the
crude prevalence of bacterial vaginosis by race or
ethnicity, education, and poverty/income ratio per-
sisted in the adjusted analyses. Black, non-Hispanic
women had an odds of bacterial vaginosis three times
that of non-Hispanic, white women after adjusting for
age, education, and poverty. Similarly, the odds for
Mexican-American women were slightly higher when
compared with white women. The adjusted analyses
confirmed associations of bacterial vaginosis with
lower levels of education and living near or below the
federal poverty level. These differences remained
with the addition of history of douching to the model.
The association with black race was attenuated some-
what (OR 2.66, 95% CI 2.18 –3.25).
In the subsample of women between the ages of
20 and 49 years, we also examined the association
with specific sexual characteristics after adjustment
for age, race or ethnicity, education, and poverty level
(data not shown). After adjustment for age, race or
ethnicity, education, and poverty level, age at first
intercourse and number of lifetime male sexual part-
ners were associated with increased odds of bacterial
vaginosis. Women who reported later ages of inter-
course had somewhat, but not significantly, lower
odds of bacterial vaginosis. Women with two to five
lifetime male sexual partners did not have a higher
odds of bacterial vaginosis (OR 1.03, 95% CI 0.69 –
1.54) than women who had one lifetime male sexual
partner, but women with six or more partners did
have a higher odds ratio (OR [6 –10 partners] 1.47,
95% CI 1.03–2.09; OR [11 or more partners] 1.62,
95% CI 1.09 –2.39). Further, having had two or more
male sexual partners in the past year was associated
with an increased odds of bacterial vaginosis com-
pared with women with a single male sexual partner
(OR 1.46, 95% CI 1.08 –1.98).
Although preliminary, these data suggest possible
interaction between race or ethnicity and other socio-
demographic characteristics. Figure 1 presents the
prevalence of bacterial vaginosis by age, education,
and poverty/income ratio for the different strata of
race or ethnicity. We sought to understand whether
the relationship between age, education, and poverty
were uniform across race or ethnic groups. For exam- Fig. 1. Prevalence of bacterial vaginosis by age (A), educa-
ple, an inspection of the association with age in the tion (B), and poverty/income ratio (C) stratified by race or
different groups indicates that bacterial vaginosis ethnicity. GED, general equivalency diploma.
Allsworth. Prevalence of Bacterial Vaginosis. Obstet Gynecol
prevalence increases with age among white, non- 2007.
Hispanic women (P⫽.33, 2 test), is flat among Mex-
118 Allsworth and Peipert Prevalence of Bacterial Vaginosis OBSTETRICS & GYNECOLOGY
ican-American women (P⫽.70), and appears to pla- ported no history of sexual intercourse had bacterial
teau among black, non-Hispanic women from 20 vaginosis. This finding confirms a previous study by
years of age and older (P⫽.21). Similarly, when Yen and colleagues,3 in which approximately 18% of
examining the association between race or ethnicity sexually inexperienced women were found to be
and poverty level, the prevalence of bacterial vagino- positive for bacterial vaginosis.
sis decreases slightly, but not significantly, within Previous studies have found that bacterial vagi-
increasing income for white and Mexican-American nosis prevalence increased with age. In a population
women (P⫽.15 and P⫽.16, respectively). Among of individuals seeking STD treatment, 23% of women
black women, there was little difference in the prev- aged 14 –24 years had bacterial vaginosis compared
alence of bacterial vaginosis among those living be- with 33% of women aged 25 years and older.17
low (55%) or near (57%) the poverty level and a Although these national data confirmed that the prev-
significant decrease in prevalence among those living alence of bacterial vaginosis is lower among women
at two times the poverty level (44%; P⫽.01). An 14 years of age, there is no evidence to support the
additional analysis, which evaluated for interactions assertion that bacterial vaginosis is associated with age
between race or ethnicity and sociodemographic among older women. This contradiction may be a
characteristics, largely confirmed these findings. The consequence of clinical study designs that dispropor-
findings were suggestive of interactions with age tionately enroll younger (and higher risk) women.
(P⫽.04), level of education (P⫽.07), and poverty/ Socioeconomic status and poverty are also asso-
income ratio (P⫽.11) and white, non-Hispanic race or ciated with the distribution of bacterial vaginosis in
ethnicity. Further, there was a significant interaction the population. Factors such as Medicaid status, low
between age and Mexican-American ethnicity. There levels of education, absence of a telephone in the
were no significant interactions with black, non-His- home, occupation, and employment status have all
panic race or ethnicity. been found to be associated with higher frequency of
bacterial vaginosis.18,20,22,23
DISCUSSION As found in several other studies,20,22,24,25 douch-
Although many studies have examined the preva- ing was associated with higher prevalence of bacterial
lence of bacterial vaginosis among different subpopu- vaginosis. Further many studies have found that the
lations, the National Health and Nutrition Examina- prevalence of douching is higher among African-
tion Survey provides an estimate of the prevalence of American populations. In this sample 15–17% of
bacterial vaginosis in the general U.S. population. white and Mexican-American women reported
The prevalence of bacterial vaginosis in multiple douching in the past 6 months compared with 44% of
studies of pregnant women ranged from 12% to black women. The addition of douching to the ad-
21%,5,15,16 was as high as 30% in people seeking health justed model resulted in a change of the effect size of
care or seeking termination of pregnancy,17,18 and was black race, but both factors remained significant
over 50% in a population of injection drug users.19 predictors of bacterial vaginosis independently. Be-
Yen and colleagues3 examined women entering the cause this analysis is cross-sectional, it is not possible
military and found that 28% of the sexually experi- to ascertain whether the association with douching is
enced and 18% of non–sexually experienced women causal or a result of attempts to self-treat vaginal
had bacterial vaginosis. Almost one third of all symptoms.
women tested positive for bacterial vaginosis. Bacterial vaginosis is common among the general
Consistent with the existing literature,18,20,21 bac- population of women in the United States. In fact, the
terial vaginosis was more common among black and prevalence of women with bacterial vaginosis in
Mexican-American women. A study in women pre- National Health and Nutrition Examination Survey
senting at a county health center in Michigan found was comparable with that in many treatment-seeking
prevalence rates of 42% among black women, 35% populations. These data confirm what has been
among Hispanic women, and 25% among white learned about the sociodemographic distribution of
women.18 The prevalences for white (23%) and His- bacterial vaginosis from clinical populations, namely,
panic (32%) women were roughly consistent with that race or ethnicity, education, and poverty are all
those in the general population but somewhat lower associated with bacterial vaginosis prevalence. One
than that observed among black women (52%). contradiction, however, is the association with age. In
Of note, these data support the conclusion that contrast with previous studies, there is not an increas-
bacterial vaginosis is not exclusively a sexually trans- ing prevalence with increasing age. Further, these
mitted condition. Almost 15% of women who re- findings indicate that the relationship between demo-
VOL. 109, NO. 1, JANUARY 2007 Allsworth and Peipert Prevalence of Bacterial Vaginosis 119
graphic characteristics and bacterial vaginosis may Nutrition Examination Analytic and Reporting Guidelines.
Hyattsville (MD): U.S. Department of Health and Human
vary by race or ethnicity. It was not clear from
Services, Centers for Disease Control and Prevention; 2005.
subgroup analyses that associations with demographic
12. Nugent RP, Krohn MA, Hillier SL. Reliability of diagnosing
characteristics were consistent across race or ethnic- bacterial vaginosis is improved by a standardized method of
ity. Bacterial vaginosis is an important predictor of gram stain interpretation. J Clin Microbiol 1991;29:297–301.
adverse reproductive outcomes, and more complete 13. Laboratory procedure manual: bacterial vaginosis . Hyattsville
understanding of the dynamics connecting these so- (MD): National Center for Health Statistics, Centers for Dis-
ease Control and Prevention; 2006.
ciodemographic characteristics will allow for the cre-
ation of targeted interventions. Additional waves of 14. MEC Laboratory component: trichomonas vaginalis and bac-
terial vaginosis. Hyattsville (MD): National Center for Health
National Health and Nutrition Examination Survey Statistics, Centers for Disease Control and Prevention; 2006.
data will be helpful in evaluating this question in 15. Hillier SL, Nugent RP, Eschenbach DA, Krohn MA, Gibbs
detail in the future. RS, Martin DH, et al. Association between bacterial vaginosis
and preterm delivery of a low-birth-weight infant. The Vaginal
REFERENCES Infections and Prematurity Study Group. N Engl J Med
1995;333:1737–42.
1. Ness RB, Kip KE, Hillier SL, Soper DE, Stamm CA, Sweet RL,
et al. A cluster analysis of bacterial vaginosis-associated micro- 16. Hay PE, Lamont RF, Taylor-Robinson D, Morgan DJ, Ison C,
flora and pelvic inflammatory disease. Am J Epidemiol 2005; Pearson J. Abnormal bacterial colonisation of the genital tract
162:585–90. and subsequent preterm delivery and late miscarriage. BMJ
1994;308:295–8.
2. Bradshaw CS, Morton AN, Garland SM, Morris MB, Moss
LM, Fairley CK. Higher-risk behavioral practices associated 17. Moi H. Prevalence of bacterial vaginosis and its association
with bacterial vaginosis compared with vaginal candidiasis. with genital infections, inflammation, and contraceptive meth-
Obstet Gynecol 2005;106:105–14. ods in women attending sexually transmitted disease and
3. Yen S, Shafer MA, Moncada J, Campbell CJ, Flinn SD, Boyer primary health clinics. Int J STD AIDS 1990;1:86–94.
CB. Bacterial vaginosis in sexually experienced and non- 18. Holzman C, Leventhal JM, Qiu H, Jones NM, Wang J. Factors
sexually experienced young women entering the military. linked to bacterial vaginosis in nonpregnant women. Am J
Obstet Gynecol 2003;102:927–33. Public Health 2001;91:1664–70.
4. Moodley P, Connolly C, Sturm AW. Interrelationships among 19. Plitt SS, Garfein RS, Gaydos CA, Strathdee SA, Sherman SG,
human immunodeficiency virus type 1 infection, bacterial Taha TE. Prevalence and correlates of chlamydia trachomatis,
vaginosis, trichomoniasis, and the presence of yeasts. J Infect neisseria gonorrhoeae, trichomonas vaginalis infections, and
Dis 2002;185:69–73. bacterial vaginosis among a cohort of young injection drug
5. Royce RA, Thorp J, Granados JL, Savitz DA. Bacterial vagi- users in Baltimore, Maryland. Sex Transm Dis 2005;32:
nosis associated with HIV infection in pregnant women from 446–53.
North Carolina. J Acquir Immune Defic Syndr Hum Retrovi- 20. Rajamanoharan S, Low N, Jones SB, Pozniak AL. Bacterial
rol 1999;20:382–6. vaginosis, ethnicity, and the use of genital cleaning agents: a
6. Cohen CR, Duerr A, Pruithithada N, Rugpao S, Hillier S, case control study. Sex Transm Dis 1999;26:404–9.
Garcia P, et al. Bacterial vaginosis and HIV seroprevalence 21. Ness RB, Hillier S, Richter HE, Soper DE, Stamm C, Bass DC,
among female commercial sex workers in Chiang Mai, Thai- et al. Can known risk factors explain racial differences in the
land. AIDS 1995;9:1093–7. occurrence of bacterial vaginosis? J Natl Med Assoc 2003;95:
7. Haggerty CL, Hillier SL, Bass DC, Ness RB. Bacterial vagino- 201–12.
sis and anaerobic bacteria are associated with endometritis. 22. Meis PJ, Goldenberg RL, Mercer BM, Iams JD, Moawad AH,
Clin Infect Dis 2004;39:990–5. Miodovnik M et al. Preterm Prediction Study: is socioeco-
8. Evans BA, Kell PD, Bond RA, MacRae KD, Slomka MJ, nomic status a risk factor for bacterial vaginosis in Black or in
Brown DW. Predictors of seropositivity to herpes simplex virus White women? Am J Perinatol 2000;17:41–5.
type 2 in women. Int J STD AIDS 2003;14:30–6. 23. Hart G. Factors associated with trichomoniasis, candidiasis and
9. Forsum U, Holst E, Larsson PG, Vasquez A, Jakobsson T, bacterial vaginosis. Int J STD AIDS 1993;4:21–5.
Mattsby-Baltzer I. Bacterial vaginosis: a microbiological and 24. Ness RB, Hillier SL, Richter HE, Soper DE, Stamm C,
immunological enigma. APMIS 2005;113:81–90. McGregor J, et al. Douching in relation to bacterial vaginosis,
10. Klebanoff MA, Schwebke JR, Zhang J, Nansel TR, Yu KF, lactobacilli, and facultative bacteria in the vagina. Obstet
Andrews WW. Vulvovaginal symptoms in women with bacte- Gynecol 2002;100:765.
rial vaginosis. Obstet Gynecol 2004;104:267–72. 25. Cottrell BH. Vaginal douching practices of women in eight
11. Centers for Disease Control and Prevention (CDC). National Florida panhandle counties. J Obstet Gynecol Neonatal Nurs
Center for Health Statistics (NCHS). National Health and 2006;35:24–33.
120 Allsworth and Peipert Prevalence of Bacterial Vaginosis OBSTETRICS & GYNECOLOGY