Bacterial vaginosis (BV) candidate bacteria: associations with BV and behavioural practices in sexually-experienced and inexperienced women

Abstract

Background: In recent years several new fastidious bacteria have been identified that display a high specificity for BV; however no previous studies have comprehensively assessed the behavioural risk associations of these bacterial vaginosis-candidate organisms (BV-COs).

Methods: We examined the associations between 8 key previously described BV-COs and BV status established by Nugent's score (NS). We also examined the sexual practices associated with each BV-CO. We incorporated 2 study populations: 193 from a sexually-inexperienced university population and 146 from a highly sexually-active clinic population. Detailed behavioural data was collected by questionnaire and vaginal smears were scored by the Nugent method. Stored samples were tested by quantitative PCR assays for the 8 BV-COs: Atopobium vaginae, Gardnerella vaginalis, Leptotrichia spp., Megasphaera type I, Sneathia spp., and the Clostridia-like bacteria BVAB1, BVAB2 and BVAB3. Associations between BV-COs and BV and behaviours were examined by univariate and multivariable analyses.

Results: On univariate analysis, all BV-COs were more common in BV compared to normal flora. However, only Megasphaera type I, BVAB2, A. vaginae and G. vaginalis were significantly independently associated with BV by multivariable analysis. Six of the eight BV-COs (Megasphaera type I, BVAB2, BVAB3, Sneathia, Leptotrichia and G. vaginalis) were rare or absent in sexually-unexposed women, and demonstrated increasing odds of detection with increasing levels of sexual activity and/or numbers of lifetime sexual partners. Only G. vaginalis and A. vaginae were commonly detected in sexually-unexposed women. Megasphaera type I was independently associated with women-who-have-sex-with women (WSW) and lifetime sexual partner numbers, while unprotected penile-vaginal-sex was associated with BVAB2 detection by multivariate analysis.

Conclusions: Four of eight key BV-COs were significantly associated with BV after adjusting for the presence of other BV-COs. The majority of BV-COs were absent or rare in sexually-unexposed women, and associated with increasing sexual exposure, suggesting potential sexual transmission of BV-COs.