Background Several bacterial candidate organisms (COs) have recently been shown to be highly specific for BV. The epidemiological profiles for these COs are unknown and no studies have examined COs in young sexually-inexperienced women, whether these COs are sexually-transmitted, or how they relate to specific sexual activities.

Methods This study incorporates two study populations: The Female University Student Study which recruited women aged 17–21 years attending the University of Melbourne, and a sexually-experienced clinic population from Melbourne Sexual Health Centre. Participants completed a questionnaire addressing demographics and detailed sexual practices. Gram-stained vaginal smears were scored by the Nugent method. Three-hundred-and-thirty-nine samples from women with normal flora and BV were selected for analysis using quantitative PCR assays (qPCR) targeting the specific 16S rRNA gene sequences of eight published COs (G vaginalis, A vaginae, Megasphaera spp., Sneathia spp., BVAB1, BVAB2, BVAB3, and Leptotrichia spp.) and L crispatus. Detection of COs and L crispatus and their total bacterial loads were compared between women with BV and normal flora. The associations between prevalence of COs and specific sexual behavioural practices were examined by univariate and multivariate analysis.

Results Analysis found all COs were strongly associated with BV compared with normal flora and L crispatus was negatively associated. G vaginalis and A vaginae were relatively common in sexually inexperienced women: however other COs were absent in a truly virginal population. When women with normal flora and BV were analysed separately, Sneathia spp., BVAB1, BVAB2, BVAB3, Leptotrichia spp. and G vaginalis all demonstrated a progressive increase in prevalence with increasing sexual experienced and increasing numbers of vaginal sexual partners see Abstract O1-S05.01 table 1. Megasphaera spp. however differed from other COs, with a higher prevalence being strongly associated with increasing oral sex frequency and oral sex partner number.

Conclusions These data provide compelling evidence for sexual transmission of several COs—with absence of COs in virginal women and increasing prevalence with increasing sexual exposure. Interestingly the COs Sneathia spp., BVAB1, BVAB2, BVAB3, Leptotrichia spp. and G vaginalis are significantly associated with vaginal sex while the epidemiological association of Megasphaera spp. differed from the other COs being significantly associated with oral sex.