Hormonal contraception: HIV transmission and progression rates

Polly Clayden, HIV i-Base

Hormonal contraception (HC) is used widely – an estimated 14 million women in sub-Saharan Africa use either injectable or oral hormonal contraception – yet the impact on HIV remains unresolved despite twenty years of research.

Two oral presentations at CROI 2012 presented new data on two aspects of possible interaction, the first from a study looking at HIV acquisition in negative women and the second disease progression in positive women using HC in Africa. [1, 2]

Sandra McCoy presented results from an evaluation of women using oral and injectable hormonal contraception in South Africa and Zimbabwe.

Dr McCoy explained that there are different possible mechanisms for increased HIV risk with HC use. Physiological and immunological changes in the genital tract of rhesus macaques in animal studies suggest the plausibility of a biological mechanism (although the extent to which data from these studies can be applied to humans remains uncertain). There also might be differences in the behavior of women who choose to use HC, for example they might be less likely to use condoms. She noted that disentangling the direct from the indirect effect is a methodological challenge

The study was an analysis using data from women aged 18 to 49 years participating in the Methods for Improving Reproductive Health in Africa (MIRA) study, a phase 3 trial of the diaphragm and lubricant gel for HIV prevention (which showed no protective effect when added to a comprehensive protection package including condoms).

MIRA participants were followed for a median of 21 months (range 12-24). They made quarterly clinic visits where they were interviewed about contraception and sexual behavior and were tested for pregnancy, HIV, and other STIs.

The investigators used Cox proportional hazards regression and marginal structural modelling to estimate the risk of HIV acquisition among non-pregnant women who reported use of combined oral contraceptive pills (COC), progestin-only pills (POP), or injectable HC (depot medroxyprogesterone acetate [DMPA] and norethisterone enantate [Net-En]) compared to women not using these methods (it was possible to differentiate between use of the two injectable methods for 86% of visits. Baseline contraceptive implant users were excluded and new users were censored at first use.

At baseline, women (n=4866) were a mean age of 29, 61% were using HC (21% COC, 14%, POP and 26% injectables). More pill users (COC 87% and POP 96%) were married than injectable contraception users (37%). Condom use was similar among HC users, 68% overall reported condom use at last sex compared to 75% of non-HC users; 26% reported always using condoms compared to 41% non HC users.

A total of 274 women seroconverted during the study, over 6913 woman years, giving a rate of 4 infections per 100 woman years.

Unadjusted and adjusted Cox models showed no association between either form of oral contraception and HIV acquisition compared to non-HC (reference): COC AHR 0.88 (95% CI 0.39-1.32), p=0.54 and POP AHR 1.02 (95% CI 0.58 – 1.81)), p=0.94. However use of any injectable contraception was associated with a 37% increased risk of HIV infection, AHR 1.37 (95% CI, 1.01 – 1.86), p=0.04. When disaggregated, neither DMPA nor Net-En significantly increased risk in the subset of women for whom these data were available, respectively, AHR 1.32 (95% CI 0.92 – 1.90), p=0.13 and 1.21 (95% CI 0.67 – 2.21). While the effect was apparent in the DMPA subgroup, it wasn’t statistically significant due to lack of power.

The investigators performed several sensitivity analyses including the potential effect of a 90 day exposure period after discontinuation of any type of HC and analyses restricted to women who reported only one type of HC or non-HC, and to women with no missed study visits. In all cases, the effects were consistent with those presented.

Using marginal structural models, direct effects analyses adjusted for dependent covariates including condom use, the risk associated with injectables remained but was attenuated (overall effect), OR 1.16 (95% CI 0.97 – 1.53). A further estimate, which mimics the effects of a highly unethical hypothetical trial with randomly assigned methods and with women constricted to use condoms infrequently or not at all (direct effect) showed an OR of 1.38 (95% CI 1.13 – 2.12).

Dr McCoy concluded that their results suggest a moderate increased risk of HIV acquisition among women using injectable contraception. The size of which was dependent of the method of analysis used.

The following presentation from Renee Heffron, showed results from an analysis of disease progression in HIV positive women receiving hormonal contraception participating in the Partners in Prevention Study Partners in Prevention – a randomised trial of acyclovir herpes suppression to reduce HIV transmission between discordant couples (there was no reduction in HIV transmission but disease progression was modestly slowed down, AHR 0.84, p=0.03)

Prospective data from 2269 women, with baseline CD4 counts >250 cells/mm3 and enrolled at 14 sites in 7 countries in East and southern Africa, were analysed to compare rates of disease progression between those using and not using HC (reference).

In this study CD4 counts were measured 6 monthly, viral load at enrollment and 6 months later and contraceptive use reported monthly using standardised questionnaires. The primary outcome was a composite endpoint of initiation of ART, CD4 decline to <200 cells/mm3 or death (not due to trauma).

Multivariate analysis was performed using adjusted Cox proportional hazards model. Time periods with IUDs and implants were excluded due to very small numbers.

At baseline, women were a median of about 30 years of age, most were married with at least one child and CD4 just below 500 cells/mm3. About 30% reported sex without a condom in the last month and the rate of pregnancy during follow up was about 20%. Of the total, 324 women used injectable, 95 oral and 1817 non-HC.

During follow up, 31.7% women reported using injectable and 12.1% oral HC at least once. Overall, 372 women experienced a disease progression event, giving a disease progression incidence of 11.5 per 100 woman years. For women using non-HC, the incidence rate was 12.3, and for those using any HC the rate was 8.54, AHR 0.74 (95% CI 0.56-0.99), p=0.04. Rates were 8.58, AHR 0.7 (95% CI 0.51-0.96), p=0.03 and 8.39 AHR 0.96 (95% CI 0.58 – 1.59), p=0.8, for the subgroups of women using injectable and oral HC respectively. Dr Heffron noted that injectable HC use was associated with a lower rate of disease progression in this analysis, but for oral HC the numbers were too small for this to be significant.

When the investigators performed sensitivity analyses assessing 6 months prior, enrollment or cumulative contraceptive use during the study they also found no increased risk for disease progression.

Among women who were negative at enrolment but acquired HIV during follow up with CD4 >500 cells/mm3 at first post seroconversion visit, the incidence rate of decline to <500 cells/mm3 was 74.4 per 100 woman years overall and 92.82 and 31.17 per 100 woman years for those using non-HC and HC respectively, AHR 0.3 (95% CI 0.07 to 1.22), p=0.09.

Dr Heffron noted that these results were reassuring with regards to HC use and disease progression.


These data add a little to the unresolved questions about hormonal contraception and HIV. WHO is currently preparing systematic reviews looking at the associations with HIV acquisition in women, HIV acquisition in men and disease progression in HIV positive women.

We recently reported that WHO upholds guidance on hormonal contraception use:

WHO recently held an expert meeting to consider the best ways to provide information to communities and health workers. We will cover this in HTB when the statements are released.


  1. McCoy S et al. Oral and injectable contraception use and risk of HIV acquisition among women: MIRA study. 19th Conference on Retroviruses and Opportunistic Infections, 5-8 March 2012, Seattle. Oral abstract 20LB.
  2. Heffron R et al. Hormonal contraceptive use and risk of HIV-1 disease progression. 19th Conference on Retroviruses and Opportunistic Infections, 5-8 March 2012, Seattle. Oral abstract 21.

Links to other websites are current at date of posting but not maintained.