Find in Library

Bacterial vaginosis (BV) is common in women who have sex with women. While cross-sectional data support a role for sexual transmission, risks for incident BV have not been prospectively studied in this group.We studied risks for BV acquisition in a prospective cohort study of women (age 16-35 years) who reported sex with other women (>or=1 partner, prior year). Women were followed for one year with examinations at quarterly visits and for genital symptoms at any time. Species-specific 16S rRNA gene PCRs for BV-associated bacteria (BVAB) were applied to vaginal fluid obtained at enrollment. Sexual behaviors were ascertained by computer-assisted interview. Of 335 participants, 239 had no BV at baseline; 199 were seen in follow-up (median follow-up 355 days, 4.0 visits/subject). Forty women experienced >or=1 BV episode. Risks for incident BV were presentation <or=14 days since onset of menses (hazard ratio (HR) 2.3 (95% CI, 1.2-4.7), report of new sex partner with BV history (HR 3.63 (1.1-11.9)), change in vaginal discharge (HR 2.6 (1.3-5.2)) and detection of any of several BVAB in vaginal fluid at enrollment, including BVAB1 (HR 6.3 (1.4-28.1)), BVAB2 (HR 18.2 (6.4-51.8)), BVAB3 (HR 12.6 (2.7-58.4)), G. vaginalis (HR 3.9 (1.5-10.4)), Atopobium vaginae (HR 4.2 (1.9-9.3)), Leptotrichia spp (9.3 (3.0-24.4)), and Megasphaera-1 (HR 11.5 (5.0-26.6)). Detection of Lactobacillus crispatus at enrollment conferred reduced risk for subsequent BV (HR 0.18 (0.08-0.4)). Detailed analysis of behavioral data suggested a direct dose-response relationship with increasing number of episodes of receptive oral-vulvovaginal sex (HR 1.02 (95% CI, 1.00-1.04).Vaginal detection of several BVAB in BV-negative women predicted subsequent BV, suggesting that changes in vaginal microbiota precede BV by weeks or months. BV acquisition was associated with report of new partner with BV; associations with sexual practices - specifically, receptive oral sex - require further investigation.