Longitudinal Change in Nugent Score and Microbiome Diversity among African Women Initiating DMPA
BM Whitney1, K Tapia1,2, S Srinivasan2, EM Muriuki3, BH Chohan1,4, JM Wallis2, C Liu2, RS McClelland1, NG Hoffman1, DN Fredricks1,2, AC Roxby1
1Univ of Washington, USA; 2Fred Hutchinson Cancer Research Center, USA; 3Univ of Nairobi, Kenya; 4Kenya Medical Research Institute, Kenya
Depot-medroxyprogesterone acetate (DMPA) has been linked to HIV acquisition, but the mechanism is unclear. DMPA reduces clinical bacterial vaginosis (BV), a known HIV risk factor, however molecular studies assessing how DMPA alters vaginal microbiota are lacking. We hypothesized that DMPA use could increase overall diversity of vaginal bacteria, therefore increasing HIV risk.
We conducted a cohort study of postpartum, breastfeeding women in Kenya initiating DMPA or non-hormonal contraception (NHC). Women received their first DMPA injection or condoms at enrollment and were followed longitudinally. Vaginal Gram stains were assessed to calculate a Nugent score for BV. Vaginal swabs were analyzed with broad-range 16S rRNA gene PCR with sequencing to assess bacterial diversity using Shannon Diversity Index (SDI). Linear mixed-effects regression adjusted for demographics, breastfeeding, sexual activity, and vaginal washing was used to generate estimates of mean changes in Nugent score and SDI over time in women using DMPA compared to those using NHC.
We enrolled 54 HIV-negative women, 33 initiating DMPA and 21 choosing NHC. At baseline, a greater proportion of DMPA users were married and had resumed sexual activity. Over 3 months, Nugent score decreased by 1.1 in NHC users (p=0.43) and 3.3 in DMPA users (p<0.001). While Nugent score decreased significantly in DMPA initiators, mean SDI did not decrease over time (=0.17, p=0.53). Mean SDI did not change significantly for women using NHC either (=-0.49; p=0.16). The aspect of Nugent score that drove the decrease over time in DMPA users was a significant decrease in small Gram-negative to -variable rods (=-2.26, p<0.001), while the other components of the score remained fairly stable.
After 3 months, DMPA users had a significant decrease in Nugent score, however they did not have a similar decrease in diversity. Higher microbiome diversity has been associated with higher HIV risk. This finding describes a possible mechanism to link DMPA with risk of HIV acquisition if women with lower Nugent scores are still colonized with key high-risk bacteria.