Risk and protective factors associated with BV chronicity among women in Rakai, Uganda

Methods A longitudinal cohort study included 255 sexually experienced, postmenarcheal women who provided weekly self-collected vaginal swabs for up to 2 years. Vaginal swabs were scored using Nugent criteria and classified as normal (≤3), intermediate (4–6) and Nugent-BV (≥7). Detailed behavioural/health information were assessed every 6 months. A per-woman longitudinal summary measure of BV chronicity was defined as the percentage of each woman’s weekly vaginal assessments scored as Nugent-BV over a 6-month interval. Risk and protective factors associated with BV chronicity were assessed using multiple linear regression with generalised estimating equations.

Results Average BV chronicity was 39% across all follow-up periods. After adjustment, factors associated with BV chronicity included baseline Nugent-BV (β=35.3, 95% CI 28.6 to 42.0) compared with normal baseline Nugent scores and use of unprotected water for bathing (ie, rainwater, pond, lake/stream) (β=12.0, 95% CI 3.4 to 20.5) compared with protected water sources (ie, well, tap, borehole). Women had fewer BV occurrences if they were currently pregnant (β=−6.6, 95% CI −12.1 to 1.1), reported consistent condom use (β=−7.7, 95% CI −14.2 to 1.3) or their partner was circumcised (β=−5.8, 95% CI −11.3 to 0.3).

Conclusions Factors associated with higher and lower values of BV chronicity were multifactorial. Notably, higher values of BV chronicity were associated with potentially contaminated bathing water. Future studies should examine the role of waterborne microbial agents in the pathogenesis of BV.