Hormonal contraception does not increase women’s HIV acquisition risk in Zambian discordant couples, 1994–2012

Abstract

Objective To determine the impact of hormonal contraceptive methods on risk of HIV acquisition among HIV-negative women cohabiting with HIV-positive male partners. Study design From 1994–2012, HIV discordant couples recruited from a couples’ voluntary HIV counseling and testing center in Lusaka, Zambia were followed longitudinally. HIV-negative partners were tested quarterly. This analysis is restricted to couples in which the man was HIV-positive and the woman was HIV-negative at enrollment and the man was not on antiretroviral treatment. Multivariate Cox models evaluated associations between time-varying contraceptive methods and HIV acquisition among women. Sensitivity analyses explored exposure misclassification and time-varying confounder mediation. Results Among 1393 couples, 252 incident infections occurred in women over 2842 couple-years (8.9 infections per 100 couple-years; 95% CI, 7.8–10.0). Multivariate Cox models indicated that neither injectable [adjusted hazard ratio (aHR)=1.2; 95% CI, 0.8–1.7], oral contraceptive pill (OCP, aHR=1.3; 95% CI, 0.9–1.8), or implant (aHR=1.1; 95% CI, 0.5–2.2) use was significantly associated with HIV acquisition relative to non-hormonal contraception controlling for woman's age, literacy and time-varying measures of genital ulceration/inflammation. This remained true when only looking at the subset of infections acquired from the spouse (82% of infections) and additionally controlling for baseline HIV viral load of the male partner, pregnancy status, and time-varying measures of sperm on a vaginal swab wet prep and self-reported unprotected sex. OCP and injectable users reported more unprotected sex (p<.001), and OCP users were more likely to have sperm on vaginal swab (p=.1) than nonhormonal method users. Conclusions We found no association between hormonal contraception and HIV acquisition risk in women. Condom use and reinforced condom counseling should always be recommended for HIV discordant couples. HIV testing of sex partners together is critical to establish HIV risk, ascertain couple fertility intentions and counsel appropriately. Implications These findings add to a controversial literature and uniquely address several common design and analytic challenges faced by previous studies. After controlling for confounders, we found no association between hormonal contraception and HIV acquisition risk in women. We support promoting condoms for HIV prevention and increasing the contraceptive method mix to decrease unintended pregnancy.

An updated systematic review of epidemiological evidence on hormonal contraceptive methods and HIV acquisition in women

“…Among 30 studies with hormonal contraceptive versus non-use of hormonal contraception comparisons, 24 included estimates specific to (or largely composed of) oral contraceptives [21–24,26,27,29,39,41–46,49–53,55–59,61,62]. Twenty-four included estimates specific to (or largely composed of) injectables [21,24–29,39–48,50–53,55,56,58,60,62] and three included implant-specific estimates [27,39,50,54].…”

“…Twenty-four included estimates specific to (or largely composed of) injectables [21,24–29,39–48,50–53,55,56,58,60,62] and three included implant-specific estimates [27,39,50,54]. All studies assessing DMPA assessed intramuscular DMPA, rather than the lower dose, subcutaneous formulation.…”

“…Of 30 hormonal contraceptive versus non-use of hormonal contraception, we rated 12 as IBWIL [21,26,27,29,39,42,43,47,51–53,55,56,58], including four newly identified studies [21,26,27,29]. Table 2 provides details on new IBWIL studies; information on previously included IBWIL studies is available elsewhere [1].…”

“…Table 2 provides details on new IBWIL studies; information on previously included IBWIL studies is available elsewhere [1]. The four new studies included a large IPD meta-analysis that assessed oral contraceptives, DMPA, and NET-EN across a range of datasets [26], an analysis from an 18-year cohort study of Zambian serodiscordant couples to assess oral contraceptives, DMPA, and implants [27], and two analyses from large microbicide trials, one assessing unspecified injectables [29] and the other assessing oral contraceptives, DMPA, and NET-EN [21]. Below, we summarize results from all 12 hormonal contraceptive versus non-use of hormonal contraception studies considered IBWIL.…”

Levonorgestrel in contraceptives and multipurpose prevention technologies

“…Only two higher-quality studies assessed levonorgestrel-containing contraceptive implants: adjHR for Norplant® [12, 25]: 1.6 (95% CI: 0.5-5.7) and adjHR for either Norplant® or Jadelle® [13]: 0.96 (95% CI: 0.29-1.34). Thus, while data are sparse and 95% CIs are wide, neither study suggested a statistically significant increased risk of HIV acquisition in women using levonorgestrel-containing implants.…”