BACTERIAL VAGINOSIS IS a common condition of poorly understood etiology in which altered vaginal flora are associated with the production of a malodorous discharge with a raised pH. Vaginal douching, particularly with antiseptic agents, may alter the vaginal milieu,1–4 increasing susceptibility to the growth of microorganisms associated with bacterial vaginosis, including Gardnerella vaginalis, Mobiluncus spp., Prevotella spp., and other anaerobes.5 Nonantiseptic preparations, such as vinegar and saline, have, however, been reported to have little effect on vaginal flora.4 In the UK, health promotion advice informs women that using vaginal douches and perfumed soaps can lead to bacterial vaginosis.6,7 However, very few studies have examined the role of vaginal douching, or of other genital cleaning practices, in the development of bacterial vaginosis, and definitions of the site of use of cleaning products or of douching are not explicit. One longitudinal study examining the protective role of hydrogen peroxide producing lactobacilli in women attending sexually transmitted diseases clinics showed “douching for cleanliness” to be an independent predictor for bacterial vaginosis.5 In a study of 26 asymptomatic health‐care workers, “use of scented soap” was more common in women with bacterial vaginosis than those with normal vaginal flora.8 Finally, among women with a history of recurrent miscarriage, “douching” with antiseptic solution (chloroxylenol 4.8%, Dettol™) was more common in women with bacterial vaginosis.9
Vaginal douching1,9,10 and bacterial vaginosis5,9,11–13 are both more common among African‐American women in the US and African Caribbean women in the UK than among women from white ethnic groups. It has been suggested that, for unknown physiologic reasons, black women have a higher vaginal pH than white women, thus predisposing them to bacterial vaginosis and trichomoniasis.14 This study and others reporting a higher prevalence of bacterial vaginosis among black women have not considered ethnic differences in genital hygiene practices as a possible confounding factor.11
We conducted a case‐control study to examine the association between bacterial vaginosis and genital hygiene practices in an ethnically mixed population of women attending a sexually transmitted diseases clinic in south London.
Subjects and Methods
Cases and controls were enrolled concurrently by SR and SBJ between October 1994 and July 1995.
Bacterial vaginosis was diagnosed microscopically from an air‐dried and Gram‐stained specimen taken by speculum examination from the lateral vaginal wall. The bacterial morphology was examined by SR and, using Nugent's method of scoring from 0 to 10, bacterial vaginosis was defined as a score of 7 or more.15–17 Absence of bacterial vaginosis was defined as a vaginal Gram‐stained specimen with a Nugent score of less than 4. Scores of 4 to 6 were excluded from the study.
The patients in the study were women aged 16 years or over with bacterial vaginosis who did not have any concurrent genital infection. Patients were enrolled at a routine 2‐week follow‐up visit if bacterial vaginosis had been diagnosed at the first attendance and if endocervical and urethral cultures for N. gonorrhoeae, vaginal culture for Candida spp., endocervical and urethral enzyme immunoassay for C. trachomatis (Syva, Palo Alto, CA), microscopy for Trichomonas vaginalis, and serological tests for syphilis were negative. The controls were women aged 16 years or over attending the 2‐week follow‐up visit who did not have bacterial vaginosis or any genital infection diagnosed.
After obtaining written consent, a structured questionnaire was administered by SR or SBJ. Information was collected about the following: age; ethnic group, self‐identified from a list of eight categories introduced in the 1991 UK Census18; occupation; smoking history; history of sexual partners in the preceding 3 months; past history of genital tract infections; and current contraceptive practice. Women were then asked about the use of soap, bubble bath, antiseptics such as Dettol™, and douching products for genital cleaning. Hand‐drawn pictures of the female genitalia were used to demonstrate the vulval mucosal area and vagina. Women were asked whether each cleaning agent was applied to the vulval mucosa (vulval use), whether it was introduced through the vaginal introitus with a finger or cloth (vaginal use), and whether it was inserted as a douche using a douching bag, syringe, or other appliance (douching).
The distribution of variables in cases and controls were described using proportions or means and standard deviations and compared using chi‐square and t tests, as appropriate. Logistic regression was used to estimate odds ratios (OR) and 95% confidence intervals (95% CI) for the individual and combined effects of demographic, contraceptive, genital hygiene, sexual behavioral factors, and the occurrence of bacterial vaginosis. Statistical analyses were conducted using STATA 5.0 (Stata Corporation, Austin, TX).
There were 100 women with bacterial vaginosis and 100 women without. Sixty women reported a previous episode of bacterial vaginosis. The occurrence of bacterial vaginosis did not appear to differ significantly across the four categories of ethnic group (Table 1). A comparison restricted to women from the two largest groups represented, however, showed that more black Caribbean than white women had bacterial vaginosis (χ2, 5.1; 1 degree of freedom (df); P = 0.02). Women in unskilled or manual occupations were significantly more likely to have bacterial vaginosis than those with nonmanual jobs.
The majority of both cases and controls reported only one sexual partner in the 3 months prior to interview. A past history of bacterial vaginosis was strongly associated with having bacterial vaginosis diagnosed at the current visit. There was a tendency for bacterial vaginosis to be more common in women who reported having had any previous uncomplicated sexually transmitted infection (gonorrhea, chlamydia, trichomoniasis, herpes, or warts) but not pelvic inflammatory disease (χ2, 0.2; 1 df; P = 0.64). The use of condoms or a diaphragm for contraception appeared to protect against bacterial vaginosis. Bacterial vaginosis was not associated with a self‐reported history of miscarriage (χ2, 0.3; 1 df; P = 0.56) or with smoking (χ2, 2.1; 1 df; P = 0.15).
Use of Genital Cleaning Agents
Most women washed the vulval area with soap, and more than a quarter of these also used it vaginally (Table 2). Soap was used equally frequently by women with and without bacterial vaginosis. Bubble bath was also commonly used, but vaginal application was infrequent. Women who put bubble bath on the vulval mucosa were twice as likely to have bacterial vaginosis as those who did not, and there was a nonstatistically significant increase in the occurrence of bacterial vaginosis among those using it vaginally. All women who used bubble bath vaginally also put it on the vulva.
Antiseptic solutions were used by only a quarter of participants. Bacterial vaginosis was over three times more common in women who used them on the vulva (OR, 3.5; 95% CI, 1.7‐6.9) and in the vagina (OR, 3.1; 95% CI, 0.6‐15.9) (Table 2). Douching was reported by only 13 women of whom 3 used a proprietary product, 2 used vinegar and water, and 8 used soap, bubble bath, or antiseptics. Bacterial vaginosis was found in 2 of the 3 women who used a proprietary product, 1 of the 2 women who used vinegar, and all 8 women who used an antiseptic or a cleaning agent. Use of any agent as a douche was strongly associated with bacterial vaginosis (OR, 6.1; 95% CI, 1.3‐28.1).
The use of soap and bubble bath was equally common among women from all ethnic groups (data not shown). Antiseptic solutions were used by only 6 of 55 (11%) white women (χ2, 11.1; 3 df; P = 0.01) compared with 38 of 111 (34%) black Caribbean, 3 of 18 (17%) black African, and 3 of 16 (19%) women in all other ethnic groups (this includes women identifying themselves as mixed race, black British, and Asian). Vaginal douching was reported by 1 of 55 (2%) white women, 10 of 111 (9%) black Caribbean, none of 18 black African women, and 2 of 16 (13%) women of other ethnic groups (χ2, 5.3; 3 df; P = 0.15).
Multivariate logistic regression models were used to further explore relationships between ethnic group, genital cleaning practices, and bacterial vaginosis (Table 3). Intravaginal use of antiseptics, bubble bath, and all other products used as a douche were combined into a single variable. Compared with white women black Caribbean women were twice as likely to have bacterial vaginosis. After controlling for differences in the use of vulval and vaginal antiseptics, bubble bath, and douches this difference was no longer statistically significant (OR, 1.6; 95% CI, 0.8‐3.3). When past history of bacterial vaginosis was also taken into account, black Caribbean women were no more likely than white women to have bacterial vaginosis (OR, 1.1; 95% CI, 0.5‐2.5). In the subgroup of 140 women enrolled who did not report a history of bacterial vaginosis before study entry, after controlling for the use of antiseptics, bubble bath, and douches, there was no association between black Caribbean ethnicity and bacterial vaginosis (OR, 1.2; 95% CI, 0.5‐3.0 for comparison with white women).
A history of bacterial vaginosis was the single factor most strongly associated with current bacterial vaginosis and, after controlling for ethnicity, use of genital cleaning agents, occupation, and barrier contraceptive use, the relationship was strengthened (OR, 13.4; 95% CI, 5.5‐32.6) (Table 4). In this full‐regression model bacterial vaginosis remained over four times more common in users of vulval antiseptics. The relationship between bacterial vaginosis and vaginal use of antiseptics, bubble bath, or douche was no longer statistically significant, although the confidence intervals are wide (OR, 2.2; 95% CI, 0.8‐5.7). Women in nonmanual occupations remained significantly less likely to have bacterial vaginosis after controlling for confounding factors.
In this case‐control study we have shown that the occurrence of bacterial vaginosis is strongly associated with the use of commercial antiseptic products. Application of these substances to the vulval mucosa, even without vaginal use, is associated with bacterial vaginosis. Ethnic differences in the use of antiseptics and douches can explain the twofold increase in the prevalence of bacterial vaginosis in black Caribbean compared with white women attending a genitourinary clinic in London.
A Causal Association?
Case‐control studies cannot establish a causal association between the use of vulval and vaginal antiseptics and douches and bacterial vaginosis: women with a vaginal discharge and a fishy odor may have used these products to get rid of the symptoms. The longitudinal study of Hawes et al.5 examined the development of bacterial vaginosis in women providing vaginal specimens over a 2‐year period. New episodes of bacterial vaginosis occurred twice as commonly (OR, 2.1; 95% CI, 1.0‐4.3) among women reporting at the baseline visit that they douched. A cross‐sectional study in the United States showed a high prevalence of vaginal douching in asymptomatic teenagers, most of whom had learned about the practice from their mothers.19 Thus, it is likely that in many instances vaginal douching precedes the onset of bacterial vaginosis.
Ethnic and Cultural Differences in Genital Hygiene Practices
Regular vaginal douching by women of reproductive age has been reported to be practiced by approximately two thirds of African‐American women, compared with one third of white American women.10 In smaller samples of military conscripts1 and teenagers,19 douching has been reported by about four fifths of black women and half of white women.19 In the UK, vaginal douching seems to be less frequent, but the population prevalence is unknown. In our study population, half of whom had bacterial vaginosis, only 7% used any cleaning agent as a douche, and 30% used any preparation intravaginally. The low prevalence of douching and intravaginal application of other cleaning products explains why strong observed associations with bacterial vaginosis were not statistically significant. Similarly, lack of power might explain the nonstatistically significant association between ethnic group and bacterial vaginosis after controlling for genital cleaning practices.
Although vaginal douching may be less common in the UK and Europe than in the United States, estimates of the prevalence of bacterial vaginosis in different groups of women are similar: 12 to 16% of pregnant women;13,16 14 to 30% of women attending sexually transmitted diseases clinics;3,20,21 and approximately 28% of women seeking termination of pregnancy.22 If genital cleaning agents are implicated in inducing changes in the vaginal flora, our findings suggest that vulval, as well as intravaginal, application may be important. In our study population most women reported washing the vulva with soap or bubble bath, and a third of black Caribbean women washed the vulva with antiseptic solution.
Studies from the United States indicate that vaginal douching is practiced more commonly by women with low educational attainment10,23 and that bacterial vaginosis is more common in women with low income.16 In our study low socioeconomic status, defined as having an unskilled or manual job, was associated with bacterial vaginosis after controlling for ethnic group and vaginal douching. However, this association may have persisted because of residual confounding due to the imprecise occupational classification.
Bacterial Vaginosis and Adverse Reproductive Tract and Pregnancy Outcome
Bacterial vaginosis has been linked to the occurrence of pelvic inflammatory disease24 and is associated with premature delivery and low birth weight.16 African‐American women are twice as likely as white women to deliver prematurely, contributing to the high rates of infant mortality experienced by black ethnic groups in the United States.11,25 In this study bacterial vaginosis was not associated with a reported history of miscarriage or pelvic inflammatory disease, but the numbers were small and self‐reports may have been unreliable. Treatment of bacterial vaginosis has been suggested as a means of reducing both the rate of preterm delivery and the ethnic disparity.11,25 Experimental studies suggest that, among women with a previous preterm birth, antibiotic treatment can reduce this risk by about a third.26,27 The contribution of behavioral factors, such as vaginal douching, to adverse pregnancy outcomes has not been assessed in prospective studies of prematurity.
The relationships among the development of bacterial vaginosis, ethnically and culturally influenced genital hygiene behaviors, and socioeconomic status are complex and poorly understood. We have shown, among our clinic population, that differences in genital cleaning practices can explain a twofold ethnic difference in the occurrence of bacterial vaginosis. Use of intravaginal cleaning agents was, however, lower than that found in US studies, so the possibility of a type 2 statistical error cannot be excluded. The importance of vulval and vaginal hygiene practices in the development of bacterial vaginosis needs to be investigated further in large longitudinal or randomized studies. Future studies of bacterial vaginosis and of associations between ethnicity and adverse pregnancy outcomes and poor reproductive health should consider the role of genital hygiene behaviors and socioeconomic factors.
1. Stock RJ, Stock ME, Hutto JM. Vaginal douching. Current concepts and practices. Obstet Gynecol 1973; 42:141–146.
2. Gresenguet G, Kreiss JK, Chapko MK, Hillier SL, Weiss NS. HIV infection and vaginal douching in central Africa. AIDS 1997; 11:101–106.
3. Moi H. Epidemiologic aspects of vaginitis and vaginosis in Scandinavia. In: Horowitz BJ, Mårdh P-A, eds. Vaginitis and Vaginosis. New York: Wiley-Liss, 1991:85–91.
4. Onderdonk AB, Delaney ML, Hinkson PL, DuBois AM. Quantitative and qualitative effects of douche preparations on vaginal microflora. Obstet Gynecol 1992; 80:333–338.
5. Hawes SE, Hillier SL, Benedetti J, et al. Hydrogen peroxide-producing lactobacilli and acquisition of vaginal infections. J Infect Dis 1996; 174:1058–1063.
6. Anonymous. Bacterial Vaginosis (77m 8/97). London: Health Education Authority, 1997.
7. Slater R. We're too clean for our own good. New Nation Mar 9, 1998; 9.
8. Priestley CJ, Jones BM, Dhar J, Goodwin L. What is normal vaginal flora? Genitourinary Med 1997; 73:23–28.
9. Llahi-Camp JM, Rai R, Ison C, Regan L, Taylor-Robinson D. Association of bacterial vaginosis with a history of second trimester miscarriage. Hum Reprod 1996; 11:1575–1578.
10. Aral SO, Mosher WD, Cates W Jr. Vaginal douching among women of reproductive age in the United States: 1988. Am J Public Health 1992; 82:210–214.
11. Fiscella K. Racial disparities in preterm births. The role of urogenital infections. Public Health Rep 1996; 111:104–113.
12. Goldenberg RL, Klebanoff MA, Nugent R, Krohn MA, Hillier S, Andrews WW. Bacterial colonization of the vagina during pregnancy in four ethnic groups. Vaginal Infections and Prematurity Study Group. Am J Obstet Gynecol 1996; 174:1618–1621.
13. Hay PE, Lamont RF, Taylor-Robinson D, Morgan DJ, Ison C, Pearson J. Abnormal bacterial colonisation of the genital tract and subsequent preterm delivery and late miscarriage. BMJ. 1994; 308:295–298.
14. Stevens Simon C, Jamison J, McGregor JA, Douglas JM. Racial variation in vaginal pH among healthy sexually active adolescents. Sex Transm Dis 1994; 21:168–172.
15. Nugent RP, Krohn MA, Hillier SL. Reliability of diagnosing bacterial vaginosis is improved by a standardized method of gram stain interpretation. J Clin Microbiol 1991; 29:297–301.
16. Hillier SL, Nugent RP, Eschenbach DA, et al. Association between bacterial vaginosis and preterm delivery of a low-birth-weight infant. The Vaginal Infections and Prematurity Study Group. N Engl J Med 1995; 333:1737–1742.
17. Rajamanoharan S, Fearn J, Low N, et al. Diagnosing bacterial vaginosis in a GU clinic: clinical versus modified Gram stain criteria. In: Presentation at the Medical Society for the Study of Venereal Diseases Meeting, January 1995, Amsterdam.
18. Office of Population Census and Surveys 1991. Census Report for Great Britian (Part A). London: HMSO, 1993.
19. Chacko MR, McGill L, Johnson TC, Smith PB, Nenney SW. Vaginal douching in teenagers attending a family planning clinic. J Adolesc Health Care 1989; 10:217–219.
20. Lossick JG. The descriptive epidemiology of vaginal trichomoniasis and bacterial vaginosis. In: Horowitz BJ, MÅrdh P-A, eds. Vaginitis and Vaginosis. New York: Wiley-Liss, 1991:77–84.
21. Nilsson U, Hellberg D, Shoubnikova M, Nilsson S, Mårdh P-A. Sexual behavior risk factors associated with bacterial vaginosis and Chlamydia trachomatis infection. Sex Transm Dis 1997; 24:241–246.
22. Blackwell AL, Thomas PD, Wareham K, Emery SJ. Health gains from screening for infection of the lower genital tract in women attending for termination of pregnancy. Lancet 1993; 342:206–210.
23. Klebanoff SJ, Hillier SL, Eschenbach DA, Waltersdorph AM. Control of the microbial flora of the vagina by H202-generating lactobacilli. J Infect Dis 1991; 164:94–100.
24. Sweet RL. Role of bacterial vaginosis in pelvic inflammatory disease. Clin Infect Dis 1995; 20(suppl 2):S271–5.
25. Holzman C, Paneth N. Preterm birth: from prediction to prevention. Am J Public Health 1998; 88:183–184.
26. Hauth JC, Goldenberg RL, Andrews WW, DuBard MB, Copper RL. Reduced incidence of preterm delivery with metronidazole and erythromycin in women with bacterial vaginosis. N Engl J Med 1995; 333:1732–1736.
27. Morales WJ, Schorr S, Albritton J. Effect of metronidazole in patients with preterm birth in preceding pregnancy and bacterial vaginosis: a placebo-controlled, double-blind study. Am J Obstet Gynecol 1994; 171:345–347.