MELBOURNE, Australia — No link between hormonal contraceptives and a woman's risk for HIV has been found after 17 years of follow-up in a Zambian cohort of HIV-discordant couples, new research shows.
However, a large meta-analysis shows an increased risk for HIV with the use of depot medroxyprogesterone acetate (Depo-Provera), but not with any other type of hormonal contraceptive.
"The effect of hormonal contraception on the risk of HIV transmission and acquisition is a debated priority public-health issue," said Kristin Wall, PhD, from the Rollins School of Public Health at Emory University in Atlanta. "These findings add to a controversial literature," she told delegates here at the 20th International AIDS Conference.
In the Zambian study, 1393 HIV-discordant couples were followed from 1995 to 2012. The male partner was HIV-positive and was not on antiretrovirals.
Data on self-reported contraceptive use and other behavioral and clinical variables were collected, and women were tested quarterly for HIV.
The hormonal methods of birth control used included implants (Norplant, Jadelle), injectable medroxyprogesterone 50 mg, and combined oral contraceptives. Nonhormonal methods included condoms, intrauterine devices, and permanent procedures.
Over 2839 couple-years, 252 HIV infections occurred.
Multivariate Cox models were used to evaluate the association between hormonal contraceptives and time to HIV infection.
Methods of contraception used at baseline were extremely well matched between the couples with seroconversion and those without.
There was no significant difference in incidence rate per 100 couple-years between nonhormonal and hormonal forms of contraception.
Table 1. Rates of Seroconversion in HIV-Negative Women in Discordant Relationships
|Contraception Used||Incidence per 100 Couple-Years||P Value|
|None or condoms||8.4||—|
For the association between injectable hormone contraception use in the previous 3 months and the risk for any HIV infection, the hazard ratio was 1.1, relative to nonhormonal contraceptive use, after variety of potentially confounding variables was controlled for.
For the use of implants, the hazard ratio was 1.0, and for the use of oral contraceptives, the hazard ratio was 1.3.
A genetic link with the male partner was established in 207 of the women who seroconverted.
For genetically linked HIV infections, there was no significant difference in HIV risk between users and nonusers of hormonal contraceptives, even when multiple confounders, including unprotected sex, circumcision status, and viral load, were controlled for, Dr. Wall reported.
"A lot of people feel that literature reports of an association between hormonal contraceptive use and HIV acquisition are confounded by unprotected sex, which we know is unreported," she told Medscape Medical News.
For example, during 60% of intervals when HIV was detected, women reported no unprotected sex in the previous 3 months. Similarly, during 40% of intervals during which pregnancy was detected, women reported no unprotected sex.
"We know that misclassification is occurring, and we were able to correct for that using sensitivity analyses. Many other studies cannot; they only collect self-reports," Dr. Wall said.
Individual Participant Data
During the same session, Charles Morrison, PhD, senior epidemiologist from FHI 360 in Durham, North Carolina, presented findings from an individual-participant meta-analysis on hormonal contraception and HIV acquisition of 18 prospective studies carried out in southern or eastern Africa.
Examining data from individual participants can help overcome some of the methodologic challenges of simply combining estimates of the effects from multiple studies, Dr. Morrison explained.
Of the individual data on 37,124 women that were pooled, there were 1830 incident HIV infections.
Medroxyprogesterone use was reported by 28% of the cohort, norethisterone enanthate (Noristerat) by 8%, some form of oral contraceptive by 19%, and no hormonal contraceptive by 43%.
The pooled adjusted hazard ratio for HIV infection was significantly higher in medroxyprogesterone users than in nonusers, but was not higher in users of any other type of injectable contraceptive or any other form of hormonal contraception.
The adjusted hazard ratio for the risk for HIV was 1.50 for medroxyprogesterone users, compared with nonhormonal contraceptive users. For users of norethisterone enanthate, the adjusted hazard ratio was 1.24.
There was no increased risk for women using oral contraceptives (adjusted hazard ratio, 1.03).
The risk for HIV was higher for users of injectable contraceptives than for users of oral contraceptives.
Table 2. Comparison of Risk With Various Forms of Contraceptives
|Contraceptives||Adjusted Hazard Ratio|
|Medroxyprogesterone vs oral contraceptives||1.43|
|Medroxyprogesterone vs norethisterone enanthate||1.32|
|Norethisterone enanthate vs oral contraceptives||1.30|
The estimated risks associated with hormonal contraceptive use were substantially lower in studies with less methodologic bias, highlighting the limitations of observational data, Dr. Morrison pointed out.
"Use of depot medroxyprogesterone, but not oral contraceptives or norethisterone enanthate, was associated with an increased risk of HIV acquisition in this large meta-analysis," Dr. Morrison concluded.
"This study underscores the need for additional safe and effective contraceptive options for women at HIV risk," he added.
In the latest recommendations from the World Health Organization (WHO), there are no restrictions on the use of any form of hormonal contraceptives for women living with or at risk for HIV. Nor does the WHO place any restrictions on progestogen-only contraceptives (pills, injectables, or implants).
Nevertheless, because the risk for HIV associated with progestogen-only injectable contraceptives remains an open question, women and couples at high risk for HIV infection should have access to other prevention measures, including male and female condoms.
Dr. Wall cautioned that neither the data from the study by her team nor the meta-analysis presented by Dr. Morrison were included in the WHO review on which the recommendations regarding the use of hormonal contraceptives were made.
"For now, I think physicians are going to adhere to these new guidelines," she told Medscape Medical News. "But they should still emphasize dual methods for high-risk women who are using progestogen-only injectables. Of course, all women at high risk for HIV should be counseled on dual-method use — meaning a modern method of contraception for unintended pregnancy prevention and condoms for HIV prevention."
The study by Dr. Wall's team was funded by a number of organizations, including the National Institutes of Child Health, the National Institute of Mental Health, and the National Institute of Allergies and Infectious Diseases. Dr. Wall and Dr. Morrison have disclosed no relevant financial relationships.
20th International AIDS Conference: Abstracts THAC0504, THAC0503, and THAC0505LB. Presented July 24, 2014.
Medscape Medical News © 2014 WebMD, LLC
Send comments and news tips to firstname.lastname@example.org.
Cite this: Hormonal Contraceptives Don't Increase HIV Risk, Study Finds - Medscape - Aug 01, 2014.