Early studies of the dynamics of HIV-1 transmission in Africa identified biological co-factors that influenced the risk of HIV-1 acquisition. Examples of this include STI, genital ulcers, and male circumcision1–3. To varying degrees, these are each host factors that can potentially be modified to reduce the risk of HIV-1 acquisition. An epidemiological study conducted among female prostitutes in Nairobi, Kenya, and published in 1991 found that oral contraceptive use was associated with an increased risk of HIV-1 infection2. Other prospective studies published between 1993 and 1996 found no association or even a decreased risk of HIV-1 infection in association with hormonal contraceptives3–5. Three studies published between 1992 and 1996 found positive associations between intramuscular depomedroxyprogesterone acetate (DMPA) and HIV-1 seropositivity or HIV-1 acquisition6–8. Therefore, at that time, there was no consensus regarding the effect of hormonal contraception and a woman's risk of acquiring HIV-1 infection. Because of the potential importance of any such association, it was recognized that there was a need to evaluate whether hormonal contraceptives were another modifiable factor that might influence HIV-1 transmission.
Hormonal contraceptives continue to be the most popular choice of fertility control for sexually active women, and any effect that they may have on HIV-1 acquisition could have a significant effect on the dynamics of the HIV epidemic. In 1993, a prospective cohort study began in female prostitutes in Mombasa, Kenya, to examine the relationship between hormonal contraception, STI, and HIV-1 acquisition9. A unique aspect of this study was the frequent follow-up schedule for women, which allowed an accurate estimation of the timing of HIV-1 acquisition and a precise measurement of hormonal contraceptive use, thereby reducing the chance of the misclassification of contraceptive exposure.
Prospective Cohort Study
We initiated a prospective cohort study in February 1993 among HIV-1-seronegative female prostitutes attending a municipal STI clinic in Mombasa, Kenya. The study was specifically designed to investigate the correlates of HIV-1 acquisition. For this reason, women were seen on a monthly basis, and contraceptive information was collected at each visit. Monthly interviews and physical examinations focused on sexual behavior, contraceptive use, and STI diagnosis and treatment. HIV serology was performed at each visit to allow a more accurate estimation of the time of HIV-1 infection. In the analysis of the relationship between hormonal contraceptive use and HIV-1 acquisition, hormonal contraceptive use was treated as a time-dependent variable. This variable was ‘on’ if a woman had used hormonal contraceptives within 115 days of a study visit. This stringent method for determining exposure to hormonal contraceptives was used to reduce the misclassification of contraceptive use and to establish a strict temporal linking of contraceptive use and HIV-1 infection. A total of 779 women who were enrolled and returned for follow-up were included in the final analysis. The median follow-up time was 224 days, and 880 PY of follow-up were accumulated. One hundred and eleven women seroconverted to HIV-1, and the median number of days between the last negative and first positive HIV test for these women was 57 days. At baseline, oral contraceptives were used by 16% of women, and 15% used DMPA.
The one-year HIV-1 incidence rate was 15% (95% CI 11–18%). In univariate analysis, women using DMPA had an HR of HIV-1 infection of 2.2 (CI 1.4–3.4) compared with women using no hormonal contraception. A trend for increased risk of HIV-1 infection was seen among OC users compared with women using no hormonal contraception (HR 1.5; CI 0.9–2.4). When oral contraceptives were divided into high-dose and low-dose pills on the basis of contraceptive steroid content, high-dose pill users had an adjusted HR for HIV-1 infection of 2.3 (CI 0.7–7.5), whereas low-dose pill users had an HR of 1.5 (CI 0.8–2.7). Multivariate analysis was conducted to control for possible confounding by demographic, sexual exposure, or biological variables. Condom use was included in multivariate models to control for its protective effect on HIV-1 acquisition. In the final model, DMPA use remained significantly associated with HIV-1 infection (HR 2.0, CI 1.3–3.1), and a trend was present for high-dose oral contraceptive use (HR 2.6, CI 0.8–8.5, P = 0.1). Findings for univariate and multivariate models are presented in Table 1.
As a result of the ongoing controversy surrounding hormonal contraceptive use and HIV-1 infection, Wang et al.10 conducted a meta-analysis of 28 published studies that examined the association between OC use and HIV-1. Heterosexual transmission was the primary mode of HIV-1 infection in all studies included in the meta-analysis. Study populations included sex workers, STI clinic attendees, discordant couples, and family planning, antenatal, HIV, and primary care clinic attendees. Sixty-eight per cent of studies reported an odds ratio (OR) of greater than one, and an overall summary OR of 1.19 (CI 0.99–1.42) was calculated.
When examined by study design, cross-sectional studies had a summary OR of 1.21 (CI 1.01–1.44) and prospective studies had an OR of 1.32 (CI 1.12–1.57). When studies were graded on the quality of measurement of contraceptive use, the highest scoring studies had a calculated OR of 1.60 (CI 1.05–2.44). When limited to studies conducted in Africa, the OR was 1.45 (CI 1.15–1.83), and the highest quality African studies had a calculated OR of 1.65 (CI 1.09–2.52). The findings from this meta-analysis are summarized in Table 2.
The results of a large, open, prospective cohort study in female prostitutes in Mombasa, Kenya, found a significant association between DMPA use and HIV-1 infection. A trend was found between the use of high-dose OCs and the acquisition of HIV-19. These associations are supported by a meta-analysis of 28 studies that found a positive association between OC use and HIV-1 acquisition.10 In addition, studies in the SIV macaque model have also demonstrated that there is an effect of progesterone on vaginal SIV acquisition11. It is interesting that in the Mombasa sex worker cohort we have observed a propensity for women to be infected by multiple diverse variants of HIV-112,13. The transmission of multiple viruses suggests increased susceptibility of the host to HIV-1 infection. It is thus noteworthy that approximately 60% of women in this cohort were infected by multiple variants, whereas none of the 10 men from Mombasa who were examined in parallel were infected by multiple viruses during heterosexual exposure13. Studies are underway to determine whether the use of hormonal contraceptives is linked to infection by a genetically heterogeneous virus population.
It is now more than 10 years since subject no. 1001 (the first subject) was enrolled in the Mombasa cohort. The first presentation of data from the Mombasa cohort at the Tenth International Conference on AIDS, in Yokohama, Japan, in 1994 was met with scepticism. Discussions at this conference demonstrated that uncertainty remains. Several other studies have found no association between hormonal contraceptives and HIV-1 risk. Rather than dismissing either conclusion as false, we should examine why diverse results have been reported. One important factor could be that the measurement of contraceptive exposure and the temporal linking of contraception and HIV infection vary across the different studies. A less accurate and less frequent measurement of exposure to contraceptives would probably result in increased misclassification bias and a tendency towards the null hypothesis of no effect. Another potential factor for variable results from different studies on hormonal contraceptive use and HIV-1 acquisition is that this variation may simply reflect real differences between risk populations. Studies of high-risk populations, such as the Mombasa sex workers, who have multiple sex partners and may be exposed to multiple HIV-1 viral variants, indicate that such groups are at increased risk of infection if they use hormonal contraceptives. However, it is possible that, for other risk groups such as monogamous women, hormonal contraceptive use does not significantly alter the risk of HIV-1 acquisition. Further studies of these different populations using frequent monitoring and accurate measures of contraceptive exposure may help to clarify any differences between risk groups and the differences seen in the results from different studies. In the interim, any fertility regulation counseling must be coupled with education about HIV and STI and condom promotion for disease prevention.
1. Cameron DW, Simonsen JN, D'Costa LJ, et al
. Female to male transmission of human immunodeficiency virus type 1: risk factors for seroconversion in men. Lancet 1989; 2:403–407.
2. Plummer FA, Simonsen JN, Cameron DW, et al
. Cofactors in male-female sexual transmission of human immunodeficiency virus type 1. J Infect Dis 1991; 163:233–239.
3. Laga M, Manoka A, Kivuvu M, et al
. Non-ulcerative sexually transmitted diseases as risk factors for HIV-1 transmission in women: results from a cohort study. AIDS 1993; 7:95–102.
4. Saracco A, Musicco M, Nicolosi A, et al
. Man-to-woman sexual transmission of HIV: a longitudinal study of 343 steady partners of infected men. J Acquir Immune Defic Syndr 1993; 6:497–502.
5. Sinei SK, Fortney JA, Kigondu CS, et al
. HIV infection in Kenyan family planning clinic attenders. Int J STD AIDS 1996; 7:65–70.
6. Bulterys M, Chao A, Habimana P, et al
. Incident HIV-1 infection in a cohort of young women in Butare, Rwanda. AIDS 1994; 8:1585–1591.
7. Rehle T, Brinkmann UK, Siraprapasiri T, et al
. Risk factors of HIV-1 infection among female prostitutes in Khon Kaen, Northeast Thailand. Infection 1992; 20:328–331.
8. Ungchusak K, Rehle T, Thammapornpilap P, et al
. Determinants of HIV infection among female commercial sex workers in northeastern Thailand: results from a longitudinal study. J Acquir Immune Defic Syndr Hum Retrovirol 1996; 12:500–507.
9. Martin HL Jr, Nyange PM, Richardson BA, et al
. Hormonal contraception, sexually transmitted diseases, and risk of heterosexual transmission of human immunodeficiency virus type 1. J Infect Dis 1998; 178:1053–1059.
10. Wang CC, Reilly M, Kreiss JK. Risk of HIV infection in oral contraceptive pill users: a meta-analysis. J Acquir Immune Defic Syndr 1999; 21:51–58.
11. Marx PA, Spira AI, Gettie A, et al
. Progesterone implants enhance SIV vaginal transmission and early virus load. Nat Med 1996; 2:1084–1089.
12. Poss M, Martin HL, Kreiss JK, et al
. Diversity in virus populations from genital secretions and peripheral blood from women recently infected with human immunodeficiency virus type 1. J Virol 1995; 69:8118–8122.
13. Long EM, Martin HR Jr, Kreiss JK, et al
. Gender differences in HIV-1 diversity at time of infection. Nat Med 2000; 6:23–25.