Article Text
Abstract
Background Gardnerella vaginalis (GV) can be detected in women with BV and without BV. Identification of four genetically distinct GV-clades (defined using clade-specific genetic markers) led to the hypothesis that there may be both pathogenic and commensal GV-clades. We conducted a study investigating the distribution and behavioural associations of GV-clades in a cohort of women-who-have-sex-with-women (WSW).
Methods Women self-collected vaginal swabs and completed behavioural questionnaires 3-monthly for 24 months or until incident BV(Nugent Score[NS]=7–10). qPCR assays were used to detect GV and the four GV-clades. Multinomial logistic regression assessed factors associated with number of GV-clades. Generalized estimating equations population-averaged models assessed factors associated with each GV-clade. Models accounted for repeated measures.
Results 369 specimens from 101 women were analysed. GV was detected in 181 specimens, and most GV-positive specimens had multiple clades present (n=119/181, 66%). Detection of multiple GV-clades was associated with smoking (adjusted relative risk ratio [RRR]:2.52; 95%CI:1.25,5.07), increased lifetime female sex partners (FSP; adjRRR:2.43; 95%CI:1.09,5.38), and a NS=4-6 (intermediate microbiota) or NS=7-10 (Nugent BV) relative to no clades. GV4 was the most prevalent clade (n=136/369; 37%; 95% CI: 32,42%), followed by GV1 (n=116/369; 31%; 95% CI: 27,36%) and GV2 (n=76/369; 21%; 95% CI: 17,25%). GV3 was uncommon (n=17/369; 5%; 95% CI:3,7%). GV1 was associated with a NS=7–10 (adjusted odds ratio[AOR]:3.87; 95%CI:1.75,8.56), smoking (AOR:2.74; 95%CI:1.28,5.87) and report of any sexual partners (AOR:3.41; 95%CI:1.18,9.86). GV2 was associated with NS=4–6 (AOR:3.28; 95%CI:1.00,10.77), sharing of sex-toys (AOR:2.30; 95%CI:1.05,5.04) and recent male sex partners (AOR:6.58; 95%CI:2.02,21.40). GV4 presence was associated with increased lifetime FSPs (AOR:3.17; 95%CI:1.25,5.07).
Conclusion GV1 and presence of multiple GV-clades was associated with Nugent BV in WSW, whereas GV2 was associated with intermediate microbiota. Individual GV-clades were associated with a range of differing sexual behaviours in adjusted analyses. These associations are of uncertain importance, but do provide support for exchange of GV-clades between sexual partners.
Disclosure No significant relationships.