INTRAVAGINAL PRACTICES ARE COMMON AMONG sexually active women and have been described in several sub-Saharan African countries, parts of Asia, and the United States.1 These practices include vaginal wiping, vaginal washing, and inserting substances into the vagina. Women report using intravaginal practices for purposes of genital hygiene, for treatment and prevention of sexually transmitted diseases (STDs), and to enhance sexual pleasure for male partners (“dry sex”).2–4 Although many women engage in intravaginal practices because of perceived health benefits, a number of studies have demonstrated potentially harmful associations with HIV-1 infection, STDs, bacterial vaginosis (BV), and obstetric and gynecologic conditions, including pelvic inflammatory disease (PID), reduced fertility, and ectopic pregnancy.5–10
Among intravaginal practices, vaginal washing has been most extensively studied as a risk factor for BV,11–13 a condition characterized by an imbalance of vaginal flora. Bacterial vaginosis is common in African women11,12 and has been associated with genital tract infections with Trichomonas vaginalis, Chlamydia trachomatis, Neisseria gonorrhoeae, PID, and also with adverse pregnancy outcomes.14–16 More recently, prospective studies have demonstrated that BV and the absence of hydrogen peroxide-producing lactobacillus species in the vagina are associated with increased risk for HIV-1 acquisition.17–19
Studies evaluating the association between vaginal washing and BV have provided conflicting results.12,20–22 The heterogeneity of the techniques, substances, and intensity of vaginal washing practices may help to explain the different findings in these studies. For example, women may perform vaginal washing by douching with a stream or jet of water, by wiping with a piece of cloth, or by using fingers to insert water, soap, or a variety of other substances in the vagina.
In contrast to vaginal washing, few studies have examined the association between BV and other intravaginal practices such as use of vaginal lubricants.23,24 Data on the association of BV with bathing are also limited.25 Understanding the associations among intravaginal practices, general hygiene, and BV is important as a means of identifying potential targets for interventions to reduce the incidence of BV. The objective of this study was to identify risk factors for BV in a cohort of Kenyan female sex workers (FSWs) who reported no current symptoms of vaginal itching or discharge.
Materials and Methods
This cross-sectional study was conducted at an outpatient communicable diseases clinic in Mombasa, Kenya. HIV-1-seronegative FSWs from an open cohort study initiated in 1993 were invited to participate in a randomized clinical trial of periodic presumptive treatment for vaginal infections. Enrollment data from trial participants were used for this analysis. Women who reported current symptoms of vaginal itching/burning or discharge were excluded from enrollment. This exclusion criterion was based on the need to provide syndromic treatment to these women rather than randomizing them to the study intervention versus placebo. Pregnant women were also excluded. This study was approved by the human subjects committees of Kenyatta National Hospital and the University of Washington. Written informed consent was obtained from all participants.
Women provided baseline data on demographic characteristics, medical and sexual history by completing a structured interview administered by trained study nurses. The enrollment interview included detailed questions about intravaginal practices during the past week, which included substances used, frequency, and timing. Vaginal washing was defined as using a finger or a piece of cloth to introduce water, soap, or other substances to clean inside the vagina. Data on bathing practices included bathing frequency and whether the pelvis was submerged into a tub or basin while bathing. Blood was collected to determine HIV-1 serostatus and a urine sample was collected to test for pregnancy. A physical examination, including a pelvic speculum examination and collection of specimens for diagnosis of genital tract infections, was performed by one of the study clinicians.
HIV-1 serostatus was determined by enzyme-linked immunosorbent assay (ELISA) (Detect-HIV; BioChem ImmunoSystems, Montreal). Positive samples were confirmed using a second ELISA (Recombigen; Cambridge Biotech, Worcester, MA). Pregnancy testing was performed using a rapid β-hCG test (Plasmatec Laboratory Products, Dorset, U.K.). A vaginal saline wet mount was examined microscopically under high power (×40) for the presence of T. vaginalis, yeast, and clue cells. A drop of 10% potassium hydroxide was then added, and the slide was examined a second time for the presence of yeast. Culture for T. vaginalis was performed in Diamond’s media. Trichomoniasis was defined as a positive culture or the presence of motile trichomonads on examination of the wet mount. Endocervical secretions were cultured on modified Thayer-Martin media for N. gonorrhoeae. Vaginal candidiasis was defined by the presence of yeast on the vaginal wet preparation or KOH preparation. Bacterial vaginosis was diagnosed by Nugent’s criteria26 with BV defined as a score of 7 to 10. Vaginal Gram stains were all evaluated by a laboratory technician who had received specialized training in this technique and was blind to the intravaginal practices reported by study participants.
We initially compared women without BV with those with BV for the covariates of interest in this study. Comparisons of bivariate outcomes were made using χ2 tests and comparisons of continuous outcomes were made using the Mann-Whitney U test. In addition, a χ2 test for trend was used to assess the relationship between the prevalence of BV and frequency of vaginal washing.
Logistic regression was used to examine the associations among intravaginal practices, bathing, and BV. The relationships between intravaginal practices and the individual microscopical components of the vaginal Gram stain score (Gram-positive rods, Gram-variable rods, and curved rods) were also examined. Because many of the women reported more than one intravaginal practice, we constructed an initial multivariate model including the main effects of interest (vaginal washing frequency, use of petroleum jelly and use of saliva as vaginal lubricants, use of cloth for cleaning inside the vagina, and infrequent bathing). This model was designed to show the independent effects of each practice. An additional multivariate model was constructed to include the main effects of interest plus known or suspected potential confounding factors of the association between intravaginal practices and BV, including age,27 sexual risk behaviors,21 and hormonal contraceptive use.28 In the models, data on the number of sex partners in the past week and the number of sex acts in the past week were dichotomized based on the median values for the study population. For all models, collinearity of the covariates was assessed using Spearman’s correlation coefficient. In no case was there an indication of collinearity of the covariates included in a model.
Between May 2003 and December 2004, 309 women were approached as potential participants in this study, of whom 237 were enrolled. Of the 72 women who did not participate, 3 reported symptoms of a vaginal infection, one was pregnant, one was prescribed metronidazole for amoebiasis, and 67 declined participation or did not return for enrollment. Compared with participants, the women who did not participate were younger (median age 27 vs. 33 years, P <0.001) and reported less use of oral contraceptive pills (20% vs. 80%, P <0.001) and depot medroxyprogesterone acetate (21% vs. 79%, P <0.001) but did not differ significantly in terms of education, workplace, marital status, sexual risk behavior (number of new partners, frequency of sex, and use of condoms), vaginal infections (trichomoniasis, candidiasis, BV), or intravaginal practices (use of finger/cloth for vaginal washing, substances used for vaginal washing, use of vaginal lubricants).
The demographic characteristics, intravaginal practices, bathing frequency, sexual risk behaviors, and prevalence of vaginal infections in the 237 enrolled women are shown in Table 1. The median age was 33 years and the median duration of education was 8 years. Two hundred six women (86.9%) reported vaginal washing, of whom 49 (23.8%) used water only, whereas 157 (76.2%) used soap or other substances (detergent = 2 [1.3%] or antiseptic = 1 [0.6%]). The most common method of vaginal washing was with a finger, which was reported by 194 (94.2%) women. Fifty-six women (23.6%) used petroleum jelly and 44 (18.6%) used saliva as a vaginal lubricant.
In this population of asymptomatic women, 80 (33.8%) were diagnosed with bacterial vaginosis, 24 (10.1%) with vaginal candidiasis, and 7 (3.0%) with trichomoniasis. Compared with women without BV, those with BV were more likely to report a new partner within the last month or report use of saliva or petroleum jelly as vaginal lubricants and less likely to have vaginal candidiasis (Table 1).
Correlates of Bacterial Vaginosis
Among 29 women who reported no vaginal washing 7 (24.1%) had BV. A higher prevalence of BV was observed among women who reported vaginal washing one to 14 times per week (24 of 76 [31.6%]), 15 to 28 times per week (34 of 100 [34.0%]), and >28 times per week (15 of 29 [51.7%]). The increase in BV with increasing vaginal washing frequency was statistically significant (χ2 test for trend, P = 0.05).
In the main-effects model examining the independent effects of the different intravaginal practices, BV was associated with vaginal washing >28 times per week, use of petroleum jelly or saliva as vaginal lubricants, and bathing <14 times per week (Table 2). The findings were similar after adjustment for potential confounding factors. The findings did not differ significantly in analyses that were stratified by the substance used for vaginal washing (water vs. soap and other substances; data not shown).
To gain a better understanding of how these practices might influence individual components of the vaginal flora, we analyzed the associations of each practice identified as a potential risk factor for BV with the bacterial morphotypes included in the vaginal Gram stain score (Table 3). These results demonstrated that different practices were associated with unique patterns of alteration of the vaginal flora identified on Gram stain. It is notable that lubrication with petroleum jelly was associated with a lower prevalence of Gram-positive rods (adjusted odds ratio [OR] = 0.4, 95% confidence interval [CI] = 0.2–0.9) and more frequent detection of Gram-variable rods (adjusted OR = 2.5, 95% CI = 1.2–4.9) and curved rods (adjusted OR = 3.1, 95% CI = 1.1–8.8), whereas lubrication with saliva was associated with increased prevalence of curved rods (adjusted OR = 3.0, 95% CI = 1.0–8.8).
This study of women without symptoms of vaginal infections identified multiple risk factors for BV in a cohort of HIV-1-seronegative FSWs. To our knowledge, this is the first study to identify lubrication with petroleum jelly or saliva as risk factors for BV. There was also a significantly increased likelihood of BV among women who bathed less frequently. A stepwise increase in the likelihood of BV was observed with increasing frequency of vaginal washing.
Although a number of studies have looked for associations between vaginal washing and BV, few have attempted to assess a dose–response effect based on the frequency of this practice.20,29 In addition, the frequency of vaginal washing was generally low in these studies. In contrast, we observed a broad range of vaginal washing practices ranging from none to over 28 times per week. The observation of a dose–response effect between vaginal washing and BV provides epidemiologic evidence of a possible causal association. In addition, the presence of a frequency-dependent effect may help to explain the heterogeneity of results from prior studies in which the frequency of vaginal washing was not specifically evaluated. Of note, although we demonstrated a significant association between vaginal washing frequency and BV, it is not clear whether these data represent a continuous stepwise effect versus a threshold effect.
To gain further insight into the biologic mechanisms through which intravaginal practices and bathing might influence the risk of BV, we examined the associations of these practices with the individual components of the vaginal Gram stain score. A healthy vaginal environment is composed primarily of Lactobacillus species, which appear as Gram-positive rods on microscopy and produce antimicrobial factors that inhibit the overgrowth of potentially pathogenic microorganisms.19,21 Bacterial vaginosis is characterized by a reduction of lactobacilli and an increased prevalence of Gram-variable rods and curved rods. We found that the use of petroleum jelly as a vaginal lubricant was associated with a reduced prevalence of lactobacilli and increased prevalence of BV-associated bacterial morphotypes. In contrast, use of saliva as a vaginal lubricant was not associated with significant changes in the prevalence of Lactobacillus morphotypes or Gram-variable rods but was associated with an increased likelihood of identifying curved rods. These findings suggest that different practices may have different effects on the individual components of the vaginal Gram stain score. Further work will be needed to replicate these findings and to understand the reasons for the differential effects of specific practices on particular bacterial morphotypes.
An interesting finding of this study was the association of less frequent bathing with an increased risk of BV. This association may initially seem counterintuitive in light of our findings that vaginal washing was associated with BV and furthermore, that most women in this population practice vaginal washing during bathing as well as at other times. Less frequent bathing could increase the risk for BV by influencing bacterial concentrations on the perineum, but it is also possible that lower bathing frequency could be a marker for other risk factors that are not measured in this study.
There were limitations to this study. First, like with other cross-sectional studies, it is not possible to definitively establish causality. Indeed, symptoms of BV could lead to vaginal washing, although this source of bias would be minimized in the current study because the women denied current symptoms of vaginal infections. Second, although our multivariate analyses adjusted for multiple potential confounding factors, some level of residual confounding may remain when adjusting for self-reported risk behaviors. Third, our analyses did not specifically evaluate differences in intravaginal practices that may occur during menses. Finally, this study was conducted in a population of FSWs who denied symptoms of vaginal itching or discharge at the time of enrollment. Although many women with BV are asymptomatic,30 and BV may increase a woman’s risk for HIV-1 infection regardless of the presence of symptoms,19 further research is needed to determine whether the risk factors for BV that we have identified in this study are generalizable to other populations.
In conclusion, this study has identified several modifiable risk factors associated with BV. Prospective studies are needed to determine whether modification of these practices reduces the risk of BV. If so, then behavioral interventions to reduce or modify vaginal washing practices could provide a simple, inexpensive, female-controlled strategy for reducing the risk of BV and its associated complications.
1. Brown JE, Brown RC. Traditional intravaginal practices and the heterosexual transmission of disease: A review. Sex Transm Dis 2000; 27:183–187.
2. Beksinska M. The practice and prevalence of dry sex among women attending a sexually transmitted diseases clinic in Lusaka, Zambia. AIDS 1999; 75:178–180.
3. Ness RB, Hillier SL, Richter HE, et al. Why women douche and why they may or may not stop. Sex Transm Dis 2003; 30:71–74.
4. Dallabetta G. Traditional vaginal agents: Use and association with HIV infection in Malawian women. AIDS 1995; 9:293–297.
5. Ankum WM, Houtzager HL, Bleker OP. Risk factors for ectopic pregnancy: A meta-analysis. Fertil Steril 1996; 65:1093–1099.
6. Jossens MO, Eskenazi B, Schachter J, Sweet RL. Risk factors for pelvic inflammatory disease. A case control study. Sex Transm Dis 1996; 23:239–247.
7. Gresenguet G, Kreiss JK, Chapko MK, Hillier SL, Weiss NS. HIV infection and vaginal douching in Central Africa. AIDS (Archive) 1996; 11:101–106.
8. Gravett MG, Nelson P, deRouen T, Critchlow C, Eschenbach DA, Holmes KK. Independent associations of bacterial vaginosis and Chlamydia trachomatis
infection with adverse pregnancy outcome. JAMA 1986; 256:1899–1903.
9. van de Wijgert JHHM, Chirenje ZM, Iliff V, et al. Effect of intravaginal practices on the vaginal and cervical mucosa of Zimbabwean women. J Acquir Immun Defic Syndr 2000; 24:62–67.
10. McClelland RS, Lavreys L, Hassan WM, Mandaliya K, Ndinya-Achola JO, Baeten JM. Vaginal washing and increased risk of HIV-1 acquisition among African women: A 10-year prospective study. AIDS 2006; 20:269–273.
11. Fonck K, Kaul R, Keli F, et al. Sexually transmitted infections and vaginal douching in a population of female sex workers in Nairobi, Kenya. Sex Transm Infect 2001; 77:271–275.
12. Demba E, Morison L, van der Loeff MS, et al. Bacterial vaginosis, vaginal flora patterns and vaginal hygiene practices in patients presenting with vaginal discharge syndrome in The Gambia, West Africa. BMC Infect Dis 2005; 5.
13. Myer L. Intravaginal practices, HIV and other sexually transmitted diseases among South African women. Sex Transm Dis 2004; 31:174–179.
14. Hawes SE, Hillier SL, Benedetti J, et al. Hydrogen peroxide producing lactobacilli and acquisition of vaginal infections. J Infect Dis 1996; 174:1058–1063.
15. Weisenfeld 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; 36:663–668.
16. Hillier SL, Nugent RP, Eschenbach DA, et al. Association between bacterial vaginosis and preterm delivery of a low birth weight infant. N Engl J Med 1995; 333:1737–1742.
17. Taha TE, Hoover DR, Dallabetta GA, et al. Bacterial vaginosis and disturbances of vaginal flora: Association with increased acquisition of HIV. AIDS 1998; 12:1699–1706.
18. Martin HL, Nyange PM, Richardson BA, et al. Hormonal contraception, sexually transmitted diseases, and risk of heterosexual transmission of human immunodeficiency virus type 1. J Infect Dis 1998; 178:1053–1059.
19. Martin HL, Richardson BA, Nyange PM, et al. Vaginal lactobacilli, microbial flora and risk of human immunodeficiency virus type-1 and sexually transmitted disease acquisition. J Infect Dis 1999; 180:1863–1868.
20. Ness RB, Hillier SL, Richter HE, et al. Douching in relation to bacterial vaginosis, lactobacilli and facultative bacteria in vagina. Obstet Gynecol 2002; 100:765–771.
21. Schwebke JR, Desmond RA, Oh MK. Predictors of bacterial vaginosis in adolescent women who douche. Sex Transm Dis 2004; 31:433–436.
22. Marrazzo JM, Koutsky LA, Eschenbach DA, Agnew K, Stine K, Hillier SL. Characterization of vaginal flora and bacterial vaginosis in women who have sex with women. J Infect Dis 2002; 185:1307–1313.
23. Bailey JV, Farquhar C, Owen C. Bacterial vaginosis in lesbians and bisexual women. Sex Transm Dis 2004; 31:691–694.
24. Nilsson U, Hellberg D, Shoubnikova M, Nilsson S, Mardh PA. Sexual behavior risk factors associated with bacterial vaginosis and Chlamydia trachomatis
infection. Sex Transm Dis 1997; 24:241–246.
25. Lowe NK, Ryan-Wenger NA. Military women’s risk factors for and symptoms of genitourinary infections during deployment. Mil Med 2003; 168:569–574.
26. 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.
27. Bukusi EA, Cohen CR, Meier AS, et al. Bacterial vaginosis: Risk factors among Kenyan women and their male partners. Sex Transm Dis 2006; 33:361–367.
28. Baeten JM, Nyange PM, Richardson BA, et al. Hormonal contraception and risk of sexually transmitted disease acquisition: Results from a prospective study. Am J Obstet Gynecol 2001; 185:380–385.
29. Rajamanoharan S, Low N, Jones SB, Pozniak AL. Bacterial vaginosis, ethnicity, and the use of genital cleaning agents: A case control study. Sex Transm Dis 1999; 26:404–409.
30. Sodhani P, Garg S, Bhalla P, Singh MM, Sharma S, Gupta S. Prevalence of bacterial vaginosis in a community setting and role of the Pap smear in its detection. Acta Cytol 2005; 49:634–638.