Female Dependent Barrier Methods

Many women are not in a position to negotiate the use of condoms by their partners. For these women, female dependent barrier methods offer some protection. Users of the diaphragm and sponge have lower rates of gonorrhea and trichomoniasis, but higher rates of candidiasis, than women whose partners use the condom (Rosenberg et al 1992) (see Figure 1). In any case, most studies show a reduced risk of all of these STDs with use of female barrier methods, when compared with use of no barrier method. The female condom is under trials, and early results indicate that it is good in preventing pregnancy as well as STDs, but the thickness of the polyurethane reduces sensitivity, and the plastic makes noise during intercourse. These issues may reduce its acceptability. Also under testing is the vaginal contraceptive film, which has viricidal properties.

Figure 1. Risk of STDs for Users of Contraceptive Sponge and Diaphragm Versus Users of Condoms

Adapted from: Rosenberg et al 1992.

Hormonal Contraceptives

There are several postulated interactions between hormonal contraceptives and STDs. There is a possible increase in chlamydia infections, and this is primarily thought to be due to the large proportion of oral contraceptive (OC) users who have ectropion. On the other hand, OC users have a reduced prevalence of PID, primarily because there is thickening of the cervical mucus, decreased menstrual flow and reduced retrograde menstruation. There has been, however, greater concern about the possible increased risk of HIV in OC users.

Possible Mechanisms Influencing HIV Transmission in OC Users

There are several possible mechanisms that influence HIV transmission in OC users:

• Cervical ectropion common in puberty, pregnancy and OC use
• Chlamydia infection more common in ectropion, and intense inflammation response possibly predisposing to HIV infection
• Altered menstrual patterns
• Local effects of progesterone 
  • thickening of cervical mucus is protective 
  • thinning of endometrial epithelial surface increases risk

The influence of cervical ectropion on HIV acquisition is shown in Table 5. The first study (Moss et al 1991) shows a significant association, but a later study by Mati et al (1995) of a larger population shows no association between ectropion and HIV. Several other studies are now being conducted that will look at this association more critically, particularly the diagnosis of ectropion.

Table 5. Cervical Ectropion and HIV

Author Odds Ratio C.I. Moss et al 1991 5.0 1.7-14.7 Mati et al 1995 1.3 0.7-2.1

In a case control study, Plummer et al (1991) showed an increased risk of HIV (RR= 4.5 [1.4–13.8]) among Nairobi prostitutes on oral contraceptives. Several issues about the study were raised, however, in that past OC use was not verified, there was selection bias with differential LFU (lost to followup) and some of the analysis was inappropriate. Subsequent studies have shown mixed results, but recent prospective studies do not show any association.

Does the use of hormonal contraceptives increase the risk of acquiring HIV infection?

In recent cross sectional studies:

• three studies show significant association, with odds ratios ranging from 1.9 to 3.9
• two studies show no association (Kapiga et al 1994; Rehle et al 1992) (see Table 6)

In recent prospective studies (see Table 7):

• three studies show no association (Laga et al 1993; Saracco et al 1993; Sinei et al 1996)

Problems with these studies included:

• The relative timing of OC use and HIV infection was not known.
• Oral contraceptive use was not validated.
• Confounders were not carefully monitored.
• There were too few OC users.

On the strength of the three prospective studies, there is no need to change contraceptive practices. Results from larger, better controlled studies, however, are eagerly awaited.

Table 6. Injectable Contraceptive Use and HIV Infection: Cross-Sectional Studies Appearing 1992–1995

Author and Year Published Population, Location Measure of Use Relative Risk 95% C. I. Mati et al 1995     FP clinic attenders, Nairobi Ever vs. not  1.4 0.9–2.1 Kapiga et al 1994     FP clinic attenders, Dar-es-Salaam Ever vs. never; Current vs. never 1.8 2.1  0.8–4.2 0.9–4.8 Rehle et al 1992     Female sex workers, NE Thailand Current vs. condoms, IUD or no method 2.9 1.0–7.9

Table 7. Oral Contraceptive Use and HIV Infection: Prospective Studies Appearing 1993–1996

Author and Year Published Population, Location Measure of Use Relative Risk 95% C. I. Laga et al 1993      Sex workers, Kinshasa Ever vs. never 0.6 0.2–2.4 Saracco et al 1993   Female partners of HIV+ men, Italy  Current vs. not  0.0 - Mati et al 1995      FP clinic attenders, Nairobi Current vs. not 0.8 0.4–1.8 Sinei et al 1996   FP clinic attenders, Nairobi  Use in last 6 months vs. not  3.5 0.8–21.5

More reassuringly, a large cross sectional study by Mati et al (1995) involving 4404 women did not show any association between current and past use of OCs, DMPA or IUD and HIV serostatus (see Figure 2). The duration of OC use was tested against HIV serostatus, and no significant association was demonstrated with up to 24 months of use (see Figure 3).

Figure 2. Contraceptive Use and HIV-1 Serostatus: Current (C) and Previous (P) Use

Adapted from: Mati et al 1995.

Figure 3. Duration of Contraceptive Use and HIV-1 Serostatus: Oral Contraceptives by Months

Adapted from: Mati et al 1995.

More

Go to Table of Contents