aDepartment of Obstetrics and Gynecology, University of Washington, Seattle, USA; bCentre for Respiratory Disease Research, Kenya Medical Research Institute, Nairobi, Kenya; and cFred Hutchinson Cancer Research Center, Seattle, USA.
Sponsorship: This study was supported by the World Health Organization (WHO 96902).
Written informed consent was obtained from each participant in the study. The protocol was approved by the Ethical Review Committee at the Kenya Medical Research Institute, Nairobi, Kenya.
Received: d: 17 April 2003; accepted: 25 June 2003.
Among women attending family planning clinics in Nairobi, Kenya, the HIV-seroprevalence rates for different contraceptive methods were: depomedroxyprogesterone acetate (DMPA) 431/3279 (13.1%), combination oral contraceptive pill 114/1073 (10.6%), and progesterone-only contraceptive pill (POCP) 45/741 (6.1%). After adjusting for age, marital status, and parity, women using the POCP had a lower HIV seroprevalence (adjusted odds ratio 0.5, 95% confidence interval 0.3–0.7) than women using DMPA. This association was most pronounced among POCP users of lower parity.
At the end of 2002, 42 million individuals were infected with HIV, and half of these infections were in women . With 65 million women using hormonal contraception worldwide, determining how sex hormones influence HIV transmission is essential . In animal models, estrogen appears to protect against SIV infection . Postmenopausal women and those using depomedroxyprogesterone acetate (DMPA) are at an increased risk of HIV acquisition, suggesting that estrogen may also protect against HIV transmission. Whereas DMPA suppresses estradiol to postmenopausal levels, the progesterone-only contraceptive pill (POCP) affects circulating estrogen levels minimally [4,5]. To date, no study has reported the relationship between the POCP and HIV infection.
Between April 2000 and June 2002, generally healthy women from five family planning clinics in Nairobi, Kenya, were tested for HIV during the recruitment of an HIV-infected cohort. Results of HIV screening served as the basis for this investigation. Approval was obtained from the Kenya Medical Research Institute Ethical Review Committee, and written informed consent was obtained from all subjects. Participants underwent a brief interview, HIV counseling and testing, and CD4 T-cell measurement for HIV- seropositive subjects. Serum was tested for HIV using a rapid method (Determine HIV-1/2; Abbott Laboratoriess, North Chicago, IL, USA) with positive results confirmed by enzyme-linked immunosorbent assay (ImmunoComb II; Orgenics, Yavne, Israel). CD4 T-lymphocytes were enumerated using FACScan (Becton Dickenson, Sunnyvale, CA, USA).
Data were analysed using SPSS for Windows 11.0 (SPSS Inc., Chicago, IL, USA). Univariate analysis was performed using chi-square for categorical, Mann–Whitney test for interval, and Student's t-test for continuous variables. Multiple linear and logistic regression were used to conduct multivariate analyses.
A total of 5301 women between the ages of 18 and 40 years were enrolled in this study. Five individuals were excluded: two lacked HIV test results and three used Norplant. Most subjects (96%) used a form of hormonal contraception: DMPA 3279 (62%); combination oral contraceptive pill (OCP) containing 30 μg ethinyl estradiol and 0.15 mg levonorgestrel 1073 (20%); and POCP containing 0.03 mg levonorgestrel 741 (14%). Participants had a median age of 24 years (range 18–40); nearly all were married (93%) and parous (99%).
A total of 627 women (12%) were HIV infected. The HIV-seroprevalence rates by reported contraceptive method were as follows: DMPA 431/3279 (13.1%); combination OCP 114/1073 (10.6%); and POCP 45/741 (6.1%). In comparison with women using DMPA, both POCP [odds ratio (OR) 0.4, 95% confidence interval (CI) 0.3–0.6] and combined OCP (OR 0.8, 95% CI 0.6–0.98) users had decreased odds of HIV infection. After performing multivariate analysis to control for age, marital status, and parity, POCP (adjusted OR 0.5, 95% CI 0.3–0.7) but not combination OCP use (adjusted OR 0.8, 95% CI 0.7–1.04) remained associated with a lower HIV seroprevalence in comparison with DMPA.
To explore further the influence of age and parity on the correlation between the hormonal contraceptive method and HIV seroprevalence, additional logistic regression models were calculated. These models were limited to the 96% of women who used hormonal contraception and who reported at least one delivery. After stratification by age, HIV-seroprevalence increased significantly with parity for those taking POCP (P = 0.03, Fig. 1a) but not for users of DMPA (Fig. 1b).
Among the 627 HIV-infected participants, the median CD4 T-cell count was 545 cells/μl (range 8–2189). The contraceptive method was not associated with the CD4 T-cell count. Each additional year of age was associated with a 20 cell/μl decline in CD4 cell count (P < 0.001). Two or more live births was associated with a 74 cell/μl higher CD4 T-cell count (P = 0.008) in comparison with one or no live births.
Our findings indicated that current POCP use was independently associated with a lower HIV seroprevalence, an effect that diminished with increasing parity. Furthermore, this study represents the largest cross-sectional investigation of the association between hormonal contraceptive methods and HIV seroprevalence, and is the first to include the POCP.
Residual confounding rather than a direct hormonal effect on HIV susceptibility may explain our results. Women choosing the POCP comprise a unique population, most of whom are newly postpartum and breastfeeding. Although our study controlled for age, parity, and marital status, other demographic and behavioral factors probably influenced the results. The fact that multivariate analysis shifted the POCP odds ratio towards one suggests that the observed ‘protective effect’ may be a result of confounding by behavioral factors. A limitation of the study was that we did not determine the breastfeeding status, a history of previous contraceptive methods, the date of the last delivery, or past and current sexual behaviors. Each of these factors could affect the risk of HIV acquisition and be related to POCP use, thereby confounding our results. Interestingly, increasing parity moderated the association of POCP with decreased HIV seroprevalence. However, without knowing more about the contraceptive history, it is difficult to interpret this finding.
Studies suggest that estrogen may protect against mucosal HIV transmission. It is thus possible that changes in estrogen levels elicited by the use of POCP and DMPA affect the susceptibility to HIV acquisition. In female macaques, none of six estrogen-treated ovariectomized animals became infected by SIV after intravaginal inoculation. In contrast, five out of six progesterone-treated macaques (83%) became infected . Estradiol levels among DMPA users are low; a recent study demonstrated that 75% of women using DMPA had estradiol levels similar to postmenopausal women . Although estradiol levels have not been well measured in women using the POCP, 40% have ovulatory patterns, 80% have menstrual cycles, and a few have an inhibition of gonadotropin release . Near physiological estrogen levels among POCP users may protect against HIV transmission. Similar studies that include more detailed contraceptive and sexual histories should be performed as a next step to confirm our observations.
1. UNAIDS. UNAIDS epidemic update: December 2002
. Geneva: World Health Organization; 2002.
2. Martin HL Jr, Nyange PM, Richardson BA, Lavreys L, Mandaliya K, Jackson DJ, et al
. Hormonal contraception, sexually transmitted diseases and risk of heterosexual transmission of human immunodeficiency virus type 1. J Infect Dis
3. Smith SM, Baskin GB, Marx PA. Estrogen protects against vaginal transmission of simian immunodeficiency virus. J Infect Dis
4. Clark MK, Sowers M, Levy BT. Magnitude and variability of sequential estradiol and progesterone concentrations in women using depot medroxyprogesterone acetate for contraception. Fertil Steril
5. Kim-Bjorklund T, Landgren BM, Johannisson E. Morphometric studies of the endometrium, the fallopian tube, and the corpus luteum during contraception with the 300 mcg norethisterone minipill. Contraception
:459–471.#m AcknowledgementsThe authors would like to thank Grace Wahome, Betty Miheso, Grace Kwanya and Lilian Makhandia for their efforts on this study, and the Director of the Kenya Medical Research Institute, for his support of the research.