Low prevalence of bacterial vaginosis in Kenyan adolescent girls and rapid incidence after first sex

Abstract

Bacterial vaginosis (BV) is a risk factor for sexually transmitted infections (STI) including HIV. Adult African women have high prevalence of BV, but it is not known when first BV occurs. We sought to describe BV in younger African women, before and after first sex, and to determine incidence of BV and significant correlates of BV incidence and recurrence. In a prospective observational cohort study enrolling adolescents with limited sexual experience, young women ages 16-21 were recruited in Thika, Kenya. Eligible participants were HIV and HSV-2 seronegative and reported zero or one lifetime sexual partner. Nugent score was determined at quarterly visits from vaginal Gram stains. Trends in BV were described over time; hazard ratios were calculated using Cox regression and relative risk of BV was estimated using generalized estimating equations and Poisson regression. We enrolled 400 participants with a median age of 18.6 years (IQR 16-21). The majority (322, 80.5%) reported no history of sex, while 78 (19.5%) reported sex with 1 partner. At enrollment, BV (Nugent score ≥7) was uncommon (21 of 375, 5.6%). Overall, 144 participants had BV at least once, for an incidence rate of 16.5 cases per 100 person-years. Prior to first sex, BV was present at 2.8% of visits, compared to 13.7% of visits post-first sex. An adjusted model of BV incidence observed that first sex was associated with more than twofold increased BV risk (aHR 2.44, p=0.009, CI 1.25-4.76). Chlamydia diagnosis (aHR 1.73, p=0.02, CI 1.1-2.8) and HSV-2 seropositivity (aHR 2.88, p=0.021, CI 1.17-7.09) were both associated with incident BV. A multivariate GEE model including all episodes of BV demonstrated risk factors including first sex, STIs, urban residence, recent sex and no income; the most important risk factor was first sex (aRR 1.92, p=0.018, CI 1.12-3.31). Probability of BV increased with each subsequent episode; mean Nugent scores increased after each BV episode. Using detailed longitudinal observation, we found that Kenyan adolescents have almost no BV prior to first sex, and that initiation of sexual activity was the strongest risk factor for both prevalent and incident BV.