Bacterial vaginosis is associated with several high-risk sexual behaviors, such as having a large number of partners, according to findings from a clinic-based Australian study.¹ Although the cause of bacterial vaginosis has not been identified, the findings support the hypothesis that the condition—which has been linked to spontaneous abortion, preterm delivery and increased vulnerability to HIV infection—may be sexually transmitted. In contrast, candidiasis, which has been associated with specific sexual practices, was not linked to high-risk sexual behaviors in this study.

The researchers studied women visiting Melbourne's primary sexual health clinic who had abnormal vaginal discharge or odor; they excluded women who were pregnant, menstruating, postmenopausal or infected with HIV, as well as those who had used lubricant or topical vaginal medication within the previous 72 hours. The researchers made "considerable effort" to offer all women with relevant symptoms the opportunity to take part in the study. Half of women in whom bacterial vaginosis was clinically diagnosed between July 2003 and August 2004 participated. All participants completed questionnaires regarding their symptoms, background characteristics and sexual behavior. They also underwent a speculum examination, during which samples of vaginal secretions were obtained for microscopy and laboratory testing.

Of the 342 women (mean age, 29) who took part in the study, 157 had bacterial vaginosis, and 51 had candidiasis, including nine who had concurrent bacterial vaginosis. In analyses of factors associated with bacterial vaginosis, the researchers classified women as either having or not having vaginosis, regardless of whether the women also had vaginal candidiasis; similarly, in analyses of risk factors for candidiasis, a woman's vaginosis status had no bearing on her candidiasis classification.

The researchers conducted univariate analyses to assess potential associations between 30 behavioral and demographic factors and bacterial vaginosis. In a multivariate analysis incorporating all factors with associations at the univariate level, bacterial vaginosis was associated with having a new sexual partner in the past year (odds ratio, 2.1), more than two male sexual partners in the last year (2.0), vaginal sex more than twice per week (2.3) and a history of trichomoniasis (4.0). Women with bacterial vaginosis also had an increased likelihood of smoking (2.1) and being Australian (1.9). Because about half of participants were using oral contraceptives, menstrual phase was not included in the main regression model; however, in a model that included days since last menses, women who used oral contraceptives had reduced odds of bacterial vaginosis (0.6).

In a multivariate analysis that compared women who were positive for Candida with those who were negative, the list of risk factors was entirely different. Candidiasis was associated with douching (odds ratio, 3.0) and a history of gonorrhea (11.2); only nine participants, however, had had gonorrhea. Women whose last menses had occurred more than 17 days earlier also had increased odds of candidiasis (2.3), although, as in the analysis of bacterial vaginosis, this factor was not included in the main regression model. The risk of candidiasis was reduced among participants who had a history of genital warts (0.3) or a new sexual partner in the past three months (0.5).

The researchers acknowledge that the study's use of clinic attendees, who may have an above-average risk of STIs, may have reduced its ability to detect relevant associations. In addition, because the study was observational, unmeasured variables may have influenced the results. Nonetheless, they observe, the findings indicate links between bacterial vaginosis and "high-risk sexual behaviors" that are associated with STIs, thus supporting "the concept that bacterial vaginosis may be sexually transmitted." (Candidiasis was also associated with sexual practices, but these behaviors were not of the high-risk variety.) Finally, the researchers point out that despite "considerable evidence" that bacterial vaginosis can be sexually transmitted between women and a substantial likelihood that it develops in similar ways in heterosexual partners, definitive proof that bacterial vaginosis is sexually transmitted will ultimately require identification of a specific causative factor.—P. Doskoch

REFERENCE

1. Bradshaw CS et al., Higher-risk behavioral practices associated with bacterial vaginosis compared with vaginal candidiasis, Obstetrics & Gynecology, 2005, 106(1):105-114.