Incident bacterial vaginosis (BV) in women who have sex with women is associated with behaviors that suggest sexual transmission of BV.

BACKGROUND Female same-sex partnerships provide a unique opportunity to study the pathogenesis and transmissibility of bacterial vaginosis (BV) because it can be diagnosed in both members of the partnership. We conducted a nationwide community-enrolled cohort study of women who have sex with women, including women coenrolled with their regular female sexual partner (FSP), to investigate the BV incidence rate and factors associated with incident BV. METHODS Women who have sex with women, without prevalent BV in a cross-sectional study, were enrolled in a 24-month cohort study involving 3-monthly questionnaires and self-collected vaginal swabs that were scored by the Nugent method. We assessed the BV incidence rate per 100 woman-years (WY) and used univariate and multivariable Cox regression analysis to establish factors associated with BV acquisition. RESULTS Two hundred ninety-eight participants were enrolled in the cohort; 122 were coenrolled with their regular FSP. There were 51 incident cases of BV (rate, 9.75/100 WY; 95% confidence interval [CI], 7.41-12.83). Incident BV was associated with exposure to a new sexual partner (adjusted hazard ratio [AHR], 2.51; 95% CI, 1.30-4.82), a partner with BV symptoms (AHR, 3.99; 95% CI, 1.39-11.45), receptive oral sex (AHR, 3.52; 95% CI, 1.41-8.79), and onset of BV symptoms (AHR, 2.80; 95% CI, 1.39-5.61). Women coenrolled with their BV-negative partner had a greatly reduced risk of incident BV (AHR, 0.26; 95% CI, .11-.61), and high concordance of Nugent category (74%), which was predominantly normal vaginal flora throughout follow-up. CONCLUSIONS These data highlight the strong influence of sexual relationships and behaviors on BV acquisition and the vaginal microbiota. They provide epidemiological evidence to support exchange of vaginal bacterial species between women and the concept that BV is sexually transmitted.