Share this article on:

Hormonal contraceptive use and female-to-male HIV transmission: a systematic review of the epidemiologic evidence

Polis, Chelsea B.; Phillips, Sharon J.; Curtis, Kathryn M.

doi: 10.1097/QAD.0b013e32835ad539
Editorial Review

Objective: To systematically review epidemiologic evidence assessing whether hormonal contraception alters the risk of HIV transmission from an HIV-positive woman to an HIV-negative male partner.

Design: Systematic review.

Methods: We included articles published or in press through December 15, 2011. We assessed studies with direct evidence on hormonal contraception use and HIV transmission, and summarized studies with indirect evidence related to genital or plasma viral load.

Results: One study provided direct evidence on oral contraceptive pills (OCPs) or injectable contraception and female-to-male HIV transmission; both injectables [Cox-adjusted hazard ratio (adjHR) 1.95, 95% confidence interval (CI) 1.06–3.58; marginal structural model (MSM) adjusted odds ratio (adjOR) 3.01, 95% CI 1.47–6.16] and OCPs (Cox adjHR 2.09, 95% CI 0.75–5.84; MSM adjOR 2.35, 95% CI 0.79–6.95) generated elevated point estimates, but only estimates for injectables were significant. Findings from 11 indirect studies assessing various hormonal contraception methods and viral genital shedding or setpoint were mixed, and seven of eight studies indicated no adverse effect of various hormonal contraception methods on plasma viral load.

Conclusion: The only direct study on OCPs or injectable contraception and female-to-male HIV transmission suggests increased risk with the use of injectables. Given the potential for confounding in observational data, the paucity of direct evidence on this subject, and mixed indirect evidence, additional evidence is needed.

aOffice of Population and Reproductive Health, United States Agency for International Development, Washington, DC, USA

bDepartment of Reproductive Health and Research, World Health Organization, Geneva, Switzerland

cDivision of Reproductive Health, Centers for Disease Control and Prevention, Atlanta, Georgia, USA.

Correspondence to Chelsea B. Polis, USAID GH/PRH/RTU, 1201 Pennsylvania Ave NW, Suite 315, Washington DC, 20004, USA. Tel: +1 202 808 3800; e-mail:

Received 13 July, 2012

Revised 18 September, 2012

Accepted 27 September, 2012

Back to Top | Article Outline


Prevention of HIV transmission (including female-to-male sexual HIV transmission) and prevention of unintended pregnancy (including unintended pregnancies to women living with HIV) are both important concerns for public health and for sexual and reproductive rights. Enabling HIV-positive women who wish to avoid pregnancy to use contraception is also a key strategy in reducing vertical HIV transmission. Hormonal contraceptives are among the most highly effective, reversible methods of pregnancy prevention. The World Health Organization has emphasized the need to understand if various methods of hormonal contraception affect HIV acquisition in HIV-negative women, HIV progression in HIV-positive women, transmission of HIV from an HIV-positive woman to an HIV-negative male sexual partner, and whether various hormonal contraceptive methods may interact with antiretroviral therapy (ART) [1].

Previous systematic reviews concluded that hormonal contraception is well tolerated by women living with HIV, but that some theoretical concerns remain [2]. Our objective was to update previous reviews on the association between hormonal contraception use and risk of HIV transmission from HIV-positive women to HIV-negative male sexual partners. In this review, we do not address HIV acquisition, progression, or drug interactions, which have been addressed elsewhere [2–5]. An earlier version of this systematic review was prepared for a WHO technical consultation in January/February 2012, during which evidence regarding the safety of hormonal contraception for women living with HIV was considered.

Several biological mechanisms could theoretically increase risk of HIV transmission from HIV-positive hormonal contraception users, but it is unclear which, if any, are relevant [6]. Higher genital HIV viral shedding has been associated with increased HIV transmission to men [7] and such shedding could be increased by direct effects of hormonal contraception on the genital tract or on local virus replication, or by indirect effects of hormonal contraception on cervical inflammation or susceptibility to sexually transmitted infections (STIs) [8]. Oral contraceptive pills (OCPs) have been associated with cervical ectopy [9,10], which has been associated with higher genital viral load in some studies [11,12] but not others [13–15]. During pregnancy, when levels of sex steroid hormones are increased, higher levels of genital HIV shedding [11,12,16–19] and higher rates of HIV transmission to men have been reported [20]. A causal relationship between hormonal contraception and various STIs remains unclear [3], but if hormonal contraception does increase STI risk, this could in turn increase genital inflammation [21], which has been associated with increased genital HIV shedding [22]. STIs such as chlamydia and gonorrhea have been associated with increased viral shedding in some studies [23], and herpes simplex virus-2 has been directly associated with increased HIV transmission [24]. Plasma viral load is a predictor of likelihood of HIV transmission [25,26], and any effect of hormonal contraception on plasma viral load could also affect infectiousness.

Back to Top | Article Outline


We searched PubMed and Embase for articles in any language ever published (or in press and brought to our attention) in a peer-reviewed journal through December 15, 2011 (search strategy available on request). We also searched reference lists of identified articles. We considered studies that compared HIV-positive women using hormonal contraception [injectables, OCPs, implants, the contraceptive patch, the contraceptive ring, or levonorgestrel intrauterine device (IUD)] against HIV-positive women not using these methods. We included randomized controlled trials (RCTs) or cohort studies with ‘direct evidence,’ in which incident HIV infection rates in male sexual partners was an outcome variable. We also included RCTs, cohort studies, or cross-sectional studies with ‘indirect evidence,’ which assessed proxy measures for infectivity in women, such as genital HIV shedding or plasma viral load, as an outcome variable.

We used Early Review Organizing Software during article selection [27]. One author conducted the literature search and identified studies for full-text review; all authors determined study inclusion and independently abstracted study information. When necessary, we attempted to contact study authors for clarifications. We did not perform meta-analysis because of limited direct evidence and heterogeneity in the design and results of studies with indirect evidence [28].

Back to Top | Article Outline

Assessment of direct evidence

By design, studies that provide direct evidence on hormonal contraception use and female-to-male HIV transmission are conducted among serodiscordant couples. We considered multiple methodologic factors that could affect study quality and results, including:

Back to Top | Article Outline

Consideration of important confounders

The potential for confounding is inherent in observational studies. Women who use hormonal contraception may differ in important ways from nonusers, and such differences could also be related to risk of HIV transmission. For example, hormonal contraception users may have higher coital frequency or less consistent condom use than nonusers of hormonal contraception [29–32], and may also differ in age, marital status, pregnancy status, or other factors.

We considered how studies assessed potential confounders, and whether they used multivariate analysis to attempt to adjust for confounding. Factors, which vary over time may result in time-dependent confounding. In such cases, marginal structural models (MSM) fit with inverse probability weights may be preferred [33–35], but these models are complex and require multiple assumptions. As with traditional statistical approaches, causal inference relies on the assumption that all confounders have been adequately measured and controlled for, or addressed with study design.

Statistical adjustment is not always sufficient to eliminate confounding. For example, information on self-reported condom use is often inaccurate [36–38], and using inadequately measured information for statistical adjustment (or failing to adjust for important covariates) can leave residual confounding. Comparing hormonal contraception users to women who use condoms as a primary contraceptive method may be problematic if consistency of condom use differs between these groups (as has been suggested in previous studies [39–42]), but is not adequately controlled. However, reasons for condom use may be unclear, and the correlation between reason for condom use and patterns of consistency of use may vary. Conducting sensitivity analyses restricted to women who report never using condoms during the study period may help to assess the robustness of results. Correlating self-reported consistent condom use with reductions in HIV or pregnancy rates, or providing information on concordance of dyad partner reports of condom use may help to increase confidence in accuracy of self-reported condom use information, and enhance confidence in successful adjustment for condom use.

Back to Top | Article Outline

Low loss to follow-up

We defined low loss to follow-up as less than 20% at 12 months.

Back to Top | Article Outline

Frequency and accuracy in measurement of exposure, outcome, and key variables

If both hormonal contraception use and HIV status are not measured repeatedly, frequently, and with respect to the same intervals of time, it is difficult to determine if hormonal contraception was used at the time of HIV transmission, or whether exposure misclassification occurred. Although nondifferential exposure misclassification could bias results towards the null, direction of the effect of differential exposure misclassification from contraceptive discontinuation, initiation, and switching on estimates is unclear, and could be more problematic for OCP versus injectable estimates, given the shorter duration of effect. Use of time-varying information, preferably in conjunction with short intersurvey intervals, can reduce misclassification. Longer intersurvey intervals increase the possibility of recall bias, make it more difficult to establish temporality, and may not accurately capture contraceptive-switching or other time-varying behaviors. We considered an intersurvey interval of less than 6 months a methodologic advantage. Most contraceptive information is collected via self-report, but validation using clinic contraceptive records can increase accuracy. Studies should collect information on type of hormonal contraception method (and dosage, if applicable), present estimates separately for different hormonal contraception methods, exclude other hormonal contraception methods from the comparison group, and use time-varying information.

Back to Top | Article Outline

Purpose of data collection

Studies specifically intended to assess the relationship between hormonal contraception use and HIV transmission may theoretically collect more comprehensive information on key variables. For secondary data analyses, the effects of study inclusion and exclusion criteria, as well as the quality of information available on key factors, should be considered. Studies should ideally specify analysis plans in advance to discourage selective reporting of statistically significant results from posthoc analyses.

Back to Top | Article Outline

Study power and precision

Studies may have limited statistical power to detect an effect if the sample size is small, the number of HIV-infected hormonal contraception users is low, or few men contract HIV. In attempting to draw causal inference, particularly in observational data, caution is justified if confidence intervals (CI) are wide and P values are marginal, particularly if point estimates are small [43].

Back to Top | Article Outline

Genetically linked transmissions

Confirming that HIV seroconversion in the male partner was genetically linked to sexual exposure from a female partner for whom hormonal contraception use status was known reduces the potential for exposure misclassification from transmission originating from a female partner in whom hormonal contraception use status is unknown.

Back to Top | Article Outline

Supplementary information on potential biological mechanisms

In studies that directly assessed hormonal contraception use and female-to-male HIV transmissions, supplementary information on hormonal contraception use and genital viral shedding is a methodologic advantage.

Back to Top | Article Outline

Assessment of indirect evidence

Studies with only indirect evidence (i.e., studies assessing the association between hormonal contraception use and plasma or genital viral load in women) did not undergo quality assessment for several reasons. First, genital viral shedding is a proxy measure for infectivity. The two metrics for assessing genital HIV viral shedding are whether virus is detected in the genital tract and the genital viral load in people with detectable virus. Evidence on the amount of increase in viral load necessary to increase risk of heterosexual HIV transmission is limited, although one study has suggested that each 1.0 log10 increase in cervical HIV RNA is significantly associated with a 67% increase in risk of HIV transmission after adjustment for plasma viral load [7]. Second, preferred methodologic approaches to assessing the association between hormonal contraception use and genital HIV viral shedding remain unclear. Multiple factors might affect the amount of virus shed in the genital tract [44], and individual day-to-day variation may also occur. Although plasma HIV viral load and genital HIV viral load are moderately correlated [44,45], differences between viral load in these compartments may arise as a consequence of genital tract inflammation or other factors [46]. Cross-sectional studies on shedding may provide a proxy for infectivity at a given point in time, but several studies have demonstrated greater within-woman variability over time of genital HIV viral load compared with plasma viral load, complicating interpretation of studies that do not use repeated measurements of genital HIV viral shedding [47,48].

Furthermore, consensus does not exist on the best way to measure genital HIV viral shedding in terms of method and site of collection [direct swab of vaginal or endocervical epithelium or cervicovaginal lavage (CVL)], type or amount of diluent used for CVL (e.g. saline, water), or outcome measured (e.g. proviral DNA, cell-free RNA, or cell-associated RNA). The preferred approaches to assess differences in genital HIV viral load between hormonal contraception users and nonusers are unknown, given the potential effects of different hormonal contraception methods on cervical mucus [49]. Ideal approaches may vary according to the characteristics of the study population (e.g. ART use) [50], and using multiple methods (e.g. swabs and CVL) may be helpful [51,52]. In terms of outcomes measured, it is unclear whether DNA, cell-free RNA, or cell-associated RNA is more indicative of infectivity [53].

For studies with indirect evidence on genital HIV viral shedding, we present selected methodologic issues, including whether studies collected genital samples at one time point (cross-sectional) or multiple time points, form of virus measured (DNA, cell-free RNA, cell-associated RNA), technique (swab, CVL, or both), site of sample collection (cervical, vaginal, or cervicovaginal), number of women assessed, whether multivariate analysis was conducted, and covariates considered. We noted whether each analysis addressed potential effects related to ART use (by restriction or statistical control), the potential for blood contamination in samples (since viral load results from a contaminated sample could be indicative of plasma HIV viral load rather than genital HIV viral load), or potential for partner contamination by screening samples for spermatozoa or Y-chromosome (samples containing seminal fluid could be indicative of the viral load in the seminal fluid of an HIV-infected male partner rather than in the female who provided the sample).

We considered genital HIV viral load as a biologic mediator between plasma HIV viral load and HIV transmission and determined it to be the more proximal indirect outcome. We briefly describe studies assessing the association between hormonal contraception and plasma HIV viral load or setpoint.

Back to Top | Article Outline


Of 634 articles collected, we assessed 23 full-text articles, excluded six for the reasons detailed in Figure 1 [54–59], and identified 17 studies eligible for inclusion. Only one had direct and indirect evidence [60]; the others had only indirect evidence on genital HIV viral shedding (in addition to information on plasma HIV viral load for some studies) [11–13,15,16,61–64], cervical HIV viral load setpoint (and plasma viral load) [17], or only plasma HIV viral load or set point [65–70]. No available studies randomized women to different hormonal contraception methods and followed them to assess differences in HIV transmission to men. Thirteen studies were conducted in sub-Saharan Africa [11–13,16,17,60,61,63,66–70], three in the United States [15,62,65], and one in Italy [64]. No studies had information about use of the contraceptive patch, contraceptive ring, or levonorgestrel IUD.

Fig. 1

Fig. 1

Back to Top | Article Outline

Direct evidence

Only one included study had direct prospective evidence on the effect of HC use on HIV transmission to men (Table 1). Heffron and colleagues conducted secondary data analysis on information from 3790 serodiscordant couples in seven African countries. Among these couples, 2476 included an HIV-positive female and HIV-negative male partner (F + M−); we focus on this subset of couples at risk of female-to-male HIV transmission. Data were primarily drawn from an RCT designed to assess the effect of acyclovir on HIV transmission, and a small proportion (about 10%) of observations were drawn from a concurrent observational study focused on determining immune correlates of HIV protection. Eligible HIV-positive female participants had no history of AIDS-defining disorders and were not using ART at enrollment. Time on ART during follow-up was censored [60]. About 11% of follow-up intervals involved self-reported sex without a condom, which was more commonly reported during intervals in which female partners reported using hormonal contraception (P = 0.009). Ninety-three incident HIV infections occurred in men (2.76 infections per 100 person-years), of which 59 (63.4%) were genetically linked to the index female partner with known hormonal contraception use status (1.75 infections per 100 person-years). Forty infections occurred in men whose partner reported not using hormonal contraception (i.e. used condoms, had a hysterectomy or tubal ligation, or used no contraception; 1.51 infections per 100 person-years) and 19 occurred to men with partners using hormonal contraception (15 used injectables and four used OCPs; 2.64 and 2.50 infections per 100 person-years, respectively).

Table 1

Table 1

Investigators used both Cox proportional hazards regression and MSM, and considered multiple covariates as potential confounders. Self-reported hormonal contraception use was included as a time-varying variable, which could be updated quarterly. The final multivariate models controlled for baseline age, baseline plasma viral load of the HIV-positive partner, time-varying pregnancy, and time-varying, self-reported unprotected sex in the last month (assumed to be representative of unprotected sex within the last 3 months). In multivariate analysis, both injectables (adjHR 1.95, 95% CI 1.06–3.58; MSM adjOR 3.01, 95% CI 1.47–6.16) and OCPs (Cox adjHR 2.09, 95% CI 0.75–5.84; MSM adjOR 2.35, 95% CI 0.79–6.95) generated elevated point estimates, but only injectables were significantly associated with HIV transmission to a male partner. Sensitivity analyses, including extending the exposure window for 3 months after last hormonal contraception use and censoring observations during pregnancy, did not result in substantial differences.

Back to Top | Article Outline

Indirect evidence

In the Heffron analysis, in addition to direct evidence, HIV-1 RNA measured 6 months after study enrollment was more likely to be detected in the genital tract of women using injectables than in women not using hormonal contraception (adjOR 1.67, 95% CI 1.21–2.31). Injectable users also had a greater concentration of genital HIV-1 RNA by an average of 0.19 log10 copies/swab (P = 0.0005) [60]. OCP use was not associated with detectability or quantity of genital HIV RNA. Hormonal contraception use and plasma HIV-1 RNA were not associated.

Nine other studies examined the influence of hormonal contraception use on genital HIV DNA or RNA shedding (expressed either as detectability or quantity) [11–13,15,16,61–64], and one study assessed various hormonal contraception methods and cervical viral load setpoint (Table 2) [17]. Most studies assessing detectability of genital viral shedding used odds ratios (OR) as a measure of association, which may exaggerate potential associations since this outcome is typically not rare (<10%) [71]. All but one study looked exclusively at HIV-1 [61]. Among 11 studies assessing genital HIV viral shedding or setpoint, seven used only a cervical or vaginal swab [11,13,16,17,60,62,63], two used only CVL [12,61], and two used a combination of swab and CVL [15,64]. Six studies were cross-sectional [11,13,15,16,60,61], whereas five collected information at multiple time points [12,17,62–64]. Six studies provided information on hormonal contraception use and plasma HIV viral load or setpoint, but did not assess hormonal contraception use and genital HIV viral shedding [65–70].

Table 2-a

Table 2-a

Table 2-b

Table 2-b

Table 2-c

Table 2-c

Back to Top | Article Outline

Genital viral detectability, quantity, or setpoint


Six studies (four of which used cross-sectional information) assessed genital HIV viral shedding with estimates specifically comparing OCP users to women not using OCPs [11,13,16,60,63,64]. Five provided estimates for cervical shedding [11,13,16,60,63], three provided estimates for vaginal shedding [11,13,63], and one provided estimates for cervicovaginal shedding [64]. Four measured DNA and showed inconsistent results: one suggested no difference in DNA detectability in the cervical compartment [16], one suggested increased DNA detectability in the cervical but not vaginal compartment [11], one suggested decreased DNA cervicovaginal detectability [64], and one showed increased DNA detectability in both the cervical and vaginal compartment, with greater increases in cervical shedding for high-dose OCPs than with low-dose OCPs [13]. Three studies measured RNA [60,63,64]: none found increased shedding in any compartment assessed, and one suggested decreased cell-associated RNA cervicovaginal detectability [64].

Back to Top | Article Outline


Three studies (two of which used cross-sectional shedding information) assessed genital viral shedding with estimates specifically comparing injectable users to women not using injectables [13,60,63]. Results of the Heffron study showed increases in cervical RNA associated with injectable use [60]. Results from the other two studies found no increase in vaginal shedding (DNA or RNA), but both reported increased cervical (DNA and RNA) shedding associated with injectable use. In one of these two studies, depot medroxyprogesterone acetate (DMPA) was associated with increased odds of cervical DNA detectability (adjOR 2.9, 95% CI 1.5–5.7) [13]; in the other study, DMPA was associated with a 0.28 log10 copy/swab increase in cervical shedding 3 months after ART initiation. This difference was not significant after controlling for plasma viral load, nor 6 months after ART initiation [63].

Back to Top | Article Outline


Only one study, which used shedding information from multiple time points, assessed genital viral shedding, with estimates specifically comparing the levonorgestrel implant (Norplant) users to women not using the implant. This study reported no change in cervical or vaginal DNA at 3 or 6 months after ART initiation [63].

Back to Top | Article Outline

Any hormonal contraception use

Four studies (two of which were cross-sectional) provided estimates on nonspecified hormonal contraception use, and none found a change in either cervical, vaginal, or cervicovaginal RNA shedding [12,15,61,62]. Combining estimates from various methods could mask effects if different methods result in effects in opposing directions.

Back to Top | Article Outline

Cervical viral load setpoint

In one study, which used shedding information at multiple time points, no association was found between use of either OCPs or DMPA and cervical viral load setpoint compared with nonusers [17].

Back to Top | Article Outline

Plasma viral load or setpoint

Nine reports on eight studies included information on plasma HIV viral load or setpoint [17,60,64–70]. One reported an association between DMPA use and higher plasma HIV viral load setpoint [70]. Although this association was not present after controlling for viral diversity [67], any change in viral load setpoint could still impact infectivity. In this study, OCPs, DMPA, and implants were not associated with a greater change in viral load over time. Six studies reported no association between various hormonal contraception methods (OCPs, injectables, or implants) and plasma RNA levels at one time point, changes in plasma RNA over time, or plasma HIV viral load setpoint [17,60,64–66,68]. One study reported an interaction between time since HIV seroconversion and use of injectable contraception: among 29 recent seroconverters, injectable users experienced a decline in plasma viral load compared with women not using injectables [69].

Back to Top | Article Outline


Direct evidence included in this review on the potential effect of OCPs or injectable contraception on the risk of female-to-male HIV transmission is limited to one study that suggests an increased risk associated with injectables. Studies with indirect evidence are heterogeneous in design and results. No studies with direct evidence included information on implants, the contraceptive patch, the contraceptive ring, or the levonorgestrel IUD, and only two studies with indirect evidence included information on implants (one related to genital viral shedding [63], the other related to plasma viral load setpoint [70]).

In addition to providing the only direct, prospectively measured estimates on use of OCPs or injectables and HIV transmission to men, the Heffron study has several methodologic strengths, including statistical adjustment for multiple potential confounders, low loss to follow-up, clear definition of exposure to hormonal contraception, use of time-varying information in conjunction with frequent follow-up, large sample size, genetic linkage of HIV transmission, and information on the use of injectables and genital viral shedding that appears to support the main analysis. However, the total number of transmissions originating from hormonal contraception users was modest (15 to partners of women using injectables and four to partners of women using OCPs), limiting statistical power and precision. Participants provided self-reported information about unprotected sex in the past month, and these responses were assumed to be accurate and consistent over the previous 3 months. Inaccuracy in measurement of this variable could have led to residual confounding in either direction.

Information on increased genital viral shedding among injectable users from the Heffron study appears to support the finding that injectable use is associated with an increase in the risk of HIV transmission. However, this evidence has some limitations. First, the amount of increased genital HIV-1 RNA shedding among injectable users was modest (+0.19 log10 copies/swab). Another analysis of these same data reported that an increase of a full 1.0 log10 copies/swab was necessary to achieve only a 67% increase in transmission risk after adjusting for plasma viral load [7], raising questions about whether the observed increase associated with injectable use was sufficient to double or triple transmission risk. Second, although point estimates for OCPs and HIV transmission to men did not reach statistical significance, they were larger than those observed for injectables, yet no increase in genital viral shedding was observed with OCPs. Multiple explanations are possible, including: OCPs may not increase risk of HIV transmission, but point estimates were elevated because of lack of precision; true differences in risk of HIV transmission between injectables and OCPs may exist, but imprecision in measurement of genital viral shedding did not reflect these differences; elevated point estimates for both OCPs and injectables were related to systematic bias related to uncontrolled confounding between hormonal contraception users and nonusers, or; both methods increase risk of HIV transmission but via different biological mechanisms. It is unclear if women with elevated genital viral loads collected 6 months after study enrollment were the same women who transmitted HIV.

Indirect evidence provides information on outcomes that are proxy measurements for infectivity, and this evidence is limited due to a lack of clarity on ideal approaches for sample collection and processing. In addition, several indirect studies were small and may have limited statistical power. Eleven indirect studies provide evidence on hormonal contraception use and genital viral shedding or setpoint, with substantial heterogeneity of methods, outcomes, and results; six studies assessed OCPs and genital viral shedding, three assessed injectables and genital viral shedding, one assessed implants and genital viral shedding, four assessed nonspecified hormonal contraception methods and genital viral shedding, and one assessed OCPs and DMPA on cervical viral load setpoint. Evidence on the association between OCPs and genital viral shedding is inconsistent. Evidence on the association between injectables and genital viral shedding is limited, mixed, and difficult to interpret. Three studies suggest increases in cervical shedding (including one for DNA [13] and two for RNA [60,63], one of which was not present after control for viral load or 3 months after ART initiation [63]), but two studies suggested no increase in vaginal shedding (one for DNA [13] and one for RNA [63]). One study found no increase in vaginal RNA shedding with use of any hormonal contraception but comprised largely of DMPA [62], and another found no increase in cervical RNA viral load setpoint with DMPA [17]. Evidence on implants and genital viral shedding is extremely limited but did not suggest an association, and evidence on other hormonal contraception methods and genital viral shedding is lacking.

Most indirect evidence indicates no association between hormonal contraception use and plasma viral load or viral load at setpoint. Plasma viral load has commonly been used as a biomarker of HIV infectivity [25], and genital viral load has been shown to predict HIV transmission to men [7]. The value of assessing cervical or plasma viral load setpoints as potential indicators of infectivity is also unknown. Although data are limited, the inconsistency in data on hormonal contraception use and genital viral load versus data on hormonal contraception use and plasma viral load may indicate that any potential effect of hormonal contraception use on viral load is localized to the genital tract, but future studies should assess if other factors, such as the potential for partner contamination in genital samples, contribute to this discrepancy.

Back to Top | Article Outline

Competing risks

HIV infection carries burdens in addition to morbidity and mortality, and HIV prevention is a public health priority. Voluntary contraceptive use offers substantial benefits to women by enabling them to prevent unintended pregnancies, which in turn decreases abortion, reduces maternal morbidity and mortality, improves child health and survival through birth spacing, and confers several nonhealth-related benefits [72–74]. HIV-positive women often report stronger desires to limit childbearing when compared with HIV-negative women, and report high levels of unmet need for contraception [75]. Effective contraception for women living with HIV who want to prevent pregnancy is also a critical and cost-effective intervention to reduce vertical HIV transmission [76–81]. Some studies suggest that pregnancy may increase the risk of genital HIV shedding [11,12,16–19] and female-to-male HIV transmission [20], but evidence is limited. Any potential effect of hormonal contraception use on female-to-male transmission must also be considered within the context of increasing access to ART, and the large reduction in risk of HIV transmission that ART use confers [82,83]. Programmatic efforts must be made to expand contraceptive method choice, as access to a wide range of highly effective contraceptive methods is limited in much of the developing world, and successful use of contraception is improved when women can obtain their method of choice.

Back to Top | Article Outline


One well conducted study raises potential concerns related to the use of injectable contraception and the risk of female-to-male HIV transmission, but given the paucity of direct evidence, mixed indirect evidence, and the potential for confounding in observational data, additional evidence is needed. Clarification on the ideal methods to assess markers of infectivity, as well as studies assessing hormonal contraception methods beyond OCPs and injectables, may also be helpful.

Back to Top | Article Outline


The authors are grateful to Mary Lyn Gaffield and Nathalie Kapp for support and guidance; Roger Chou, Mark Helfand, and Ann Duerr for expert advice; Agustin Ciapponi and Demián Glujovsky for support with EROS software; and Nellie Kamau and LaToya Armstrong for assistance with our search strategy. A subgroup of the advisory committee for the WHO Hormonal Contraception and HIV consultation, including Tsungai Chipato, Anna Glasier, Ron Gray, Ying-Ru Jacqueline Lo, and Nguyen-Toan Tran, provided valuable comments on the article. We are also grateful to authors who provided additional information on their analyses. C.B.P. conducted the literature search and identified studies for full-text review. C.B.P., S.J.P., and K.M.C. assessed included studies and wrote the manuscript.

Disclosure of financial sources: There were no external sources of support. USAID, WHO, and CDC contributed staff time to conduct this review.

Disclaimer: The findings and conclusions in this article are those of the authors and do not necessarily represent the official position of the United States Agency for International Development, the World Health Organization, or the Centers for Disease Control and Prevention.

Back to Top | Article Outline

Conflicts of interest

C.B.P. is collaborating on a study assessing the acceptability of two types of injectable contraceptive methods among HIV-positive women in Uganda (PI: Ron Gray); one of the products for that study was donated by Pfizer. S.J.P. and K.M.C. have no conflicts of interest.

Back to Top | Article Outline


1. World Health Organization. Review of Priorities in Research on Hormonal Contraception and IUDs and HIV Infection: Report of a Technical Meeting. 13–15 March 2007. Geneva, Switzerland.
2. Curtis KM, Nanda K, Kapp N. Safety of hormonal and intrauterine methods of contraception for women with HIV/AIDS: a systematic review. AIDS 2009; 23:S55–S67.
3. Morrison CS, Turner AN, Jones LB. Highly effective contraception and acquisition of HIV and other sexually transmitted infections. Best Pract Res Clin Obstet Gynaecol 2009; 23:263–284.
4. El-Ibiary SY, Cocohoba JM. Effects of HIV antiretrovirals on the pharmacokinetics of hormonal contraceptives. Eur J Contracept Reprod Healthcare 2008; 13:123–132.
5. Kancheva Landolt NT, Lakhonphon S, Ananworanich J. Contraception in HIV-positive female adolescents. AIDS Res Ther 2011; 8:19.
6. Blish CA, Baeten JM. Hormonal contraception and HIV-1 transmission. Am J Reprod Immunol 2011; 65:302–307.
7. Baeten JM, Kahle E, Lingappa JR, Coombs RW, Delany-Moretlwe S, Nakku-Joloba E, et al. Genital HIV-1 RNA predicts risk of heterosexual HIV-1 transmission. Sci Transl Med 2011; 3:77ra29.
8. Hel Z, Stringer E, Mestecky J. Sex steroid hormones, hormonal contraception, and the immunobiology of human immunodeficiency virus-1 infection. Endocr Rev 2010; 31:79–97.
9. Mati JK, Hunter DJ, Maggwa BN, Tukei PM. Contraceptive use and the risk of HIV infection in Nairobi, Kenya. Int J Gynaecol Obstet 1995; 48:61–67.
10. Critchlow CW, Wolner-Hanssen P, Eschenbach DA, Kiviat NB, Koutsky LA, Stevens CE, et al. Determinants of cervical ectopia and of cervicitis: age, oral contraception, specific cervical infection, smoking, and douching. Am J Obstet Gynecol 1995; 173:534–543.
11. Clemetson DB, Moss GB, Willerford DM, Hensel M, Emonyi W, Holmes KK, et al. Detection of HIV DNA in cervical and vaginal secretions. Prevalence and correlates among women in Nairobi, Kenya. JAMA 1993; 269:2860–2864.
12. Tanton C, Weiss HA, Le Goff J, Changalucha J, Rusizoka M, Baisley K, et al. Correlates of HIV-1 genital shedding in Tanzanian women. PLoS One 2011; 6:e17480.
13. Mostad SB, Overbaugh J, DeVange DM, Welch MJ, Chohan B, Mandaliya K, et al. Hormonal contraception, vitamin A deficiency, and other risk factors for shedding of HIV-1 infected cells from the cervix and vagina. Lancet 1997; 350:922–927.
14. Ghys PD, Fransen K, Diallo MO, Ettiegne-Traore V, Coulibaly IM, Yeboue KM, et al. The associations between cervicovaginal HIV shedding, sexually transmitted diseases and immunosuppression in female sex workers in Abidjan, Cote d’Ivoire. AIDS 1997; 11:F85–F93.
15. Kovacs A, Wasserman SS, Burns D, Wright DJ, Cohn J, Landay A, et al. Determinants of HIV-1 shedding in the genital tract of women. Lancet 2001; 358:1593–1601.
16. Kreiss J, Willerford DM, Hensel M, Emonyi W, Plummer F, Ndinya-Achola J, et al. Association between cervical inflammation and cervical shedding of human immunodeficiency virus DNA. J Infect Dis 1994; 170:1597–1601.
17. Morrison CS, Demers K, Kwok C, Bulime S, Rinaldi A, Munjoma M, et al. Plasma and cervical viral loads among Ugandan and Zimbabwean women during acute and early HIV-1 infection. AIDS 2010; 24:573–582.
18. Henin Y, Mandelbrot L, Henrion R, Pradinaud R, Coulaud JP, Montagnier L. Virus excretion in the cervicovaginal secretions of pregnant and nonpregnant HIV-infected women. J Acquir Immune Defic Syndr 1993; 6:72–75.
19. Gardella B, Roccio M, Maccabruni A, Mariani B, Panzeri L, Zara F, et al. HIV shedding in cervico-vaginal secretions in pregnant women. Curr HIV Res 2011; 9:313–320.
20. Mugo NR, Heffron R, Donnell D, Wald A, Were EO, Rees H, et al. Increased risk of HIV-1 transmission in pregnancy: a prospective study among African HIV-1-serodiscordant couples. AIDS 2011; 25:1887–1895.
21. Ghanem KG, Shah N, Klein RS, Mayer KH, Sobel JD, Warren DL, et al. Influence of sex hormones, HIV status, and concomitant sexually transmitted infection on cervicovaginal inflammation. J Infect Dis 2005; 191:358–366.
22. Johnson LF, Lewis DA. The effect of genital tract infections on HIV-1 shedding in the genital tract: a systematic review and meta-analysis. Sex Transm Dis 2008; 35:946–959.
23. Rotchford K, Strum AW, Wilkinson D. Effect of coinfection with STDs and of STD treatment on HIV shedding in genital-tract secretions: systematic review and data synthesis. Sex Transm Dis 2000; 27:243–248.
24. Ward H, Ronn M. Contribution of sexually transmitted infections to the sexual transmission of HIV. Curr Opin HIV AIDS 2010; 5:305–310.
25. Quinn TC, Wawer MJ, Sewankambo N, Serwadda D, Li C, Wabwire-Mangen F, et al. Viral load and heterosexual transmission of human immunodeficiency virus type 1. Rakai Project Study Group. N Engl J Med 2000; 342:921–929.
26. Attia S, Egger M, Muller M, Zwahlen M, Low N. Sexual transmission of HIV according to viral load and antiretroviral therapy: systematic review and meta-analysis. AIDS 2009; 23:1397–1404.
27. Glujovsky D, Bardach A, García Martí S, Comandé D, Ciapponi A. EROS: A new software for early stage of systematic reviews. Value in Health 2011; 14:A564.
28. Egger M, Smith GD, Schneider M. Systematic reviews of observational studies. In: Egger M, Smith GD, Altman DG, editors. Systematic reviews in healthcare: meta-analysis in context. London: BMJ Publishing Group; 2001. pp. 211–227.
29. Cushman LF, Romero D, Kalmuss D, Davidson AR, Heartwell S, Rulin M. Condom use among women choosing long-term hormonal contraception. Fam Plann Persp 1998; 30:240–243.
30. Sangi-Haghpeykar H, Posner SF, Poindexter AN 3rd. Consistency of condom use among low-income hormonal contraceptive users. Perspect Sex Reprod Health 2005; 37:184–191.
31. Polis CB, Gray RH, Lutalo T, Nalugoda F, Kagaayi J, Kigozi G, et al. Trends and correlates of hormonal contraceptive use among HIV-infected women in Rakai, Uganda, 1994–2006. Contraception 2011; 83:549–555.
32. Morrison CS, Richardson BA, Mmiro F, Chipato T, Celentano DD, Luoto J, et al. Hormonal contraception and the risk of HIV acquisition. AIDS 2007; 21:85–95.
33. Robins JM, Hernan MA, Brumback B. Marginal structural models and causal inference in epidemiology. Epidemiology 2000; 11:550–560.
34. Hernan MA, Brumback B, Robins JM. Marginal structural models to estimate the causal effect of zidovudine on the survival of HIV-positive men. Epidemiology 2000; 11:561–570.
35. Cole SR, Hernan MA, Robins JM, Anastos K, Chmiel J, Detels R, et al. Effect of highly active antiretroviral therapy on time to acquired immunodeficiency syndrome or death using marginal structural models. Am J Epidemiol 2003; 158:687–694.
36. Gallo MF, Sobel JD, Rompalo AM, Cu-Uvin S, Schoenbaum E, Jamieson DJ. Discordance between spermatozoa detection and self-reported semen exposure. Sex Transm Dis 2011; 38:909–912.
37. Minnis AM, Steiner MJ, Gallo MF, Warner L, Hobbs MM, van der Straten A, et al. Biomarker validation of reports of recent sexual activity: results of a randomized controlled study in Zimbabwe. Am J Epidemiol 2009; 170:918–924.
38. Catania JA, Gibson DR, Chitwood DD, Coates TJ. Methodological problems in AIDS behavioral research: influences on measurement error and participation bias in studies of sexual behavior. Psychol Bull 1990; 108:339–362.
39. Magwali TL, Steiner MJ, Toms H, Brown JM. How are condoms used in a family planning setting: evidence from Zimbabwe. Cent Afr J Med 2005; 51:79–84.
40. Aklilu M, Messele T, Tsegaye A, Biru T, Mariam DH, van Benthem B, et al. Factors associated with HIV-1 infection among sex workers of Addis Ababa, Ethiopia. AIDS 2001; 15:87–96.
41. Callegari L, Harper CC, van der Straten A, Kamba M, Chipato T, Padian NS. Consistent condom use in married Zimbabwean women after a condom intervention. Sex Transm Dis 2008; 35:624–630.
42. Van Rossem R, Meekers D, Akinyemi Z. Consistent condom use with different types of partners: evidence from two Nigerian surveys. AIDS Educ Prev 2001; 13:252–267.
43. Goodman SN. Toward evidence-based medical statistics. 1: The P value fallacy. Ann Intern Med 1999; 130:995–1004.
44. Anderson BL, Cu-Uvin S. Determinants of HIV shedding in the lower genital tract of women. Curr Infect Dis Rep 2008; 10:505–511.
45. Coombs RW, Reichelderfer PS, Landay AL. Recent observations on HIV type-1 infection in the genital tract of men and women. AIDS 2003; 17:455–480.
46. Cu-Uvin S, Delong AK, Venkatesh KK, Hogan JW, Ingersoll J, Kurpewski J, et al. Genital tract HIV-1 RNA shedding among women with below detectable plasma viral load. AIDS 2010; 24:2489–2497.
47. Larke NL, Weiss HA, Mayaud P, Van de Perre P, Clayton T, Ouedraogo A, et al. Design of epidemiological studies measuring genital and plasma HIV-1 outcomes: lessons from a randomised controlled trial. Trop Med Int Health 2009; 14:267–275.
48. Coombs RW, Wright DJ, Reichelderfer PS, Burns DN, Cohn J, Cu-Uvin S, et al. Variation of human immunodeficiency virus type 1 viral RNA levels in the female genital tract: implications for applying measurements to individual women. J Infect Dis 2001; 184:1187–1191.
49. Duerr A, Curtis K, Shelton JD, Meirik O. Hormonal contraception and genital-tract shedding of HIV-1-infected cells. Lancet 1998; 351:294–295.
50. Mitchell C, Paul K, Agnew K, Gaussman R, Coombs RW, Hitti J. Estimating volume of cervicovaginal secretions in cervicovaginal lavage fluid collected for measurement of genital HIV-1 RNA levels in women. J Clin Microbiol 2011; 49:735–736.
51. Dezzutti CS, Hendrix CW, Marrazzo JM, Pan Z, Wang L, Louissaint N, et al. Performance of swabs, lavage, and diluents to quantify biomarkers of female genital tract soluble mucosal mediators. PLoS One 2011; 6:e23136.
52. Delany S, Rosas R, Mlaba N, Clayton T, Akpomiemie G, LeGoff J, et al. Comparison of cervicovaginal lavage, cervicovaginal lavage enriched with cervical swab, and vaginal tampon for the detection of HIV-1 RNA and HSV-2 DNA in genital secretions. J Acquir Immune Defic Syndr 2008; 49:406–409.
53. Baeten JM, Overbaugh J. Measuring the infectiousness of persons with HIV-1: opportunities for preventing sexual HIV-1 transmission. Current HIV Res 2003; 1:69–86.
54. de Vincenzi I. A longitudinal study of human immunodeficiency virus transmission by heterosexual partners. European Study Group on Heterosexual Transmission of HIV. N Engl J Med 1994; 331:341–346.
55. European Study Group on Heterosexual Transmission of HIV. Comparison of female to male and male to female transmission of HIV in 563 stable couples. European Study Group on Heterosexual Transmission of HIV. BMJ 1992; 304:809–813.
56. Wang CC, McClelland RS, Overbaugh J, Reilly M, Panteleeff DD, Mandaliya K, et al. The effect of hormonal contraception on genital tract shedding of HIV-1. AIDS 2004; 18:205–209.
57. Heikinheimo O, Lehtovirta P, Suni J, Paavonen J. The levonorgestrel-releasing intrauterine system (LNG-IUS) in HIV-infected women--effects on bleeding patterns, ovarian function and genital shedding of HIV. Hum Reprod 2006; 21:2857–2861.
58. Heikinheimo O, Lehtovirta P, Aho I, Ristola M, Paavonen J. The levonorgestrel-releasing intrauterine system in human immunodeficiency virus-infected women: a 5-year follow-up study. Am J Obstet Gynecol 2011; 204:126e121–126e124.
59. Richardson BA, Morrison CS, Sekadde-Kigondu C, Sinei SK, Overbaugh J, Panteleeff DD, et al. Effect of intrauterine device use on cervical shedding of HIV-1 DNA. AIDS 1999; 13:2091–2097.
60. Heffron R, Donnell D, Rees H, Celum C, Mugo N, Were E, et al. Use of hormonal contraceptives and risk of HIV-1 transmission: a prospective cohort study. Lancet Infect Dis 2012; 12:19–26.
61. Seck K, Samb N, Tempesta S, Mulanga-Kabeya C, Henzel D, Sow PS, et al. Prevalence and risk factors of cervicovaginal HIV shedding among HIV-1 and HIV-2 infected women in Dakar, Senegal. Sex Transm Infect 2001; 77:190–193.
62. Clark RA, Theall KP, Amedee AM, Dumestre J, Wenthold L, Kissinger PJ. Lack of association between genital tract HIV-1 RNA shedding and hormonal contraceptive use in a cohort of Louisiana women. Sex Transm Dis 2007; 34:870–872.
63. Graham SM, Masese L, Gitau R, Jalalian-Lechak Z, Richardson BA, Peshu N, et al. Antiretroviral adherence and development of drug resistance are the strongest predictors of genital HIV-1 shedding among women initiating treatment. J Infect Dis 2010; 202:1538–1542.
64. Roccio M, Gardella B, Maserati R, Zara F, Iacobone D, Spinillo A. Low-dose combined oral contraceptive and cervicovaginal shedding of human immunodeficiency virus. Contraception 2011; 83:564–570.
65. Cejtin HE, Lisa JB, Springer C, Watts DH, Levine A, Greenblatt R, et al. Effect of hormonal contraceptive use on plasma HIV-1-RNA levels among HIV-infected women. AIDS 2003; 17:1702–1704.
66. Richardson BA, Otieno PA, Mbori-Ngacha D, Overbaugh J, Farquhar C, John-Stewart GC. Hormonal contraception and HIV-1 disease progression among postpartum Kenyan women. AIDS 2007; 21:749–753.
67. Sagar M, Lavreys L, Baeten JM, Richardson BA, Mandaliya K, Ndinya-Achola JO, et al. Identification of modifiable factors that affect the genetic diversity of the transmitted HIV-1 population. AIDS 2004; 18:615–619.
68. Polis CB, Gray RH, Bwanika JB, Kigozi G, Kiwanuka N, Nalugoda F, et al. Effect of hormonal contraceptive use before HIV seroconversion on viral load setpoint among women in Rakai, Uganda. J Acquir Immune Defic Syndr 2011; 56:125–130.
69. Kumwenda JJ, Makanani B, Taulo F, Nkhoma C, Kafulafula G, Li Q, et al. Natural history and risk factors associated with early and established HIV type 1 infection among reproductive-age women in Malawi. Clin Infect Dis 2008; 46:1913–1920.
70. Lavreys L, Baeten JM, Kreiss JK, Richardson BA, Chohan BH, Hassan W, et al. Injectable contraceptive use and genital ulcer disease during the early phase of HIV-1 infection increase plasma virus load in women. J Infect Dis 2004; 189:303–311.
71. McNutt LA, Wu C, Xue X, Hafner JP. Estimating the relative risk in cohort studies and clinical trials of common outcomes. Am J Epidemiol 2003; 157:940–943.
72. Singh SD, J.E. Adding it up: costs and benefits of contraceptive services - Estimates for 2012. New York: Guttmacher Institute and United Nations Population Fund (UNFPA); 2012.
73. Shulman LP. The state of hormonal contraception today: benefits and risks of hormonal contraceptives: combined estrogen and progestin contraceptives. Am J Obstet Gynecol 2011; 205:S9–13.
74. Burke AE. The state of hormonal contraception today: benefits and risks of hormonal contraceptives: progestin-only contraceptives. Am J Obstet Gynecol 2011; 205:S14–S17.
75. Johnson KB, Akwara P, Rutstein SO, Bernstein S. Fertility preferences and the need for contraception among women living with HIV: the basis for a joint action agenda. AIDS 2009; 23 (Suppl 1):S7–S17.
76. Reynolds HW, Janowitz B, Homan R, Johnson L. The value of contraception to prevent perinatal HIV transmission. Sex Transm Dis 2006; 33:350–356.
77. Reynolds HW, Janowitz B, Wilcher R, Cates W. Contraception to prevent HIV-positive births: current contribution and potential cost savings in PEPFAR countries. Sex Transm Infect 2008; 84 (Suppl 2):ii49–ii53.
78. Reynolds HW, Steiner MJ, Cates W Jr. Contraception's proved potential to fight HIV. Sex Trans Infect 2005; 81:184–185.
79. Halperin DT, Stover J, Reynolds HW. Benefits and costs of expanding access to family planning programs to women living with HIV. AIDS 2009; 23 (Suppl 1):S123–S130.
80. Hladik W, Stover J, Esiru G, Harper M, Tappero J. The contribution of family planning towards the prevention of vertical HIV transmission in Uganda. PLoS One 2009; 4:e7691.
81. Mahy M, Stover J, Kiragu K, Hayashi C, Akwara P, Luo C, et al. What will it take to achieve virtual elimination of mother-to-child transmission of HIV? An assessment of current progress and future needs. Sex Transm Infect 2010; 86 (Suppl 2):ii48–ii55.
82. Cohen MS, Chen YQ, McCauley M, Gamble T, Hosseinipour MC, Kumarasamy N, et al. Prevention of HIV-1 infection with early antiretroviral therapy. N Engl J Med 2011; 365:493–505.
83. Anglemyer A, Rutherford GW, Baggaley RC, Egger M, Siegfried N. Antiretroviral therapy for prevention of HIV transmission in HIV-discordant couples. Cochrane Database Syst Rev 2011: CD009153.

contraception; female-to-male HIV transmission; HIV; hormonal; prevention of mother-to-child transmission

Copyright © 2013 Wolters Kluwer Health, Inc.