Gonorrhea as a risk factor for HIV acquisition

Abstract

Objective: The role of gonorrhea in facilitating acquisition of HIV infection has only recently been studied. A previous nested case-control analysis in a cohort of female sex workers in Zaïre found a strong association between HIV seroconversion and prior gonorrheal infection. The objective of this study was to replicate the Zaïre study analysis in a cohort of 273 Cameroonian sex workers to determine whether gonorrhea increased the risk of HIV acquisition, and if the crude association between gonorrheal infection and HIV acquisition was weakened when the level of unprotected coitus was more carefully controlled.

Methods: We conducted a nested case-control study of initially HIV-1-negative women (n = 273) followed prospectively (with monthly sexually transmitted disease check-ups and 3-monthly HIV-1 serology). As in Zaïre, cases (seroconverters, n = 17) were compared with controls (women who remained HIV-1-negative, n = 68) for incidence of gonorrhea and sexual exposure during the presumed period of HIV-1 acquisition.

Results: The association between gonorrheal infection and subsequent HIV acquisition was stronger in Zaïre than in Cameroon [crude odds ratios (OR), 6.3 versus 2.2]. In both the Zaïre and Cameroon data the crude OR were reduced (6.3 to 4.8, and 2.2 to 1.7, respectively) by controlling for risk factors including a dichotomous variable indicating irregular or no condom use. When this variable was replaced in the Cameroon data with a more precise continuous variable indicating the percentage of unprotected coital acts, the gonorrhea OR was further reduced to 1.4 (95% confidence interval, 0.4-4.9).

Conclusion: These results suggest that in the Cameroon cohort, gonorrheal infection did not facilitate HIV acquisition, but that having gonorrhea was a marker for unprotected coitus that facilitated HIV acquisition. The data demonstrate how OR can be overestimated when imprecise dichotomous measures of unprotected coitus are used. Future studies should plan for better control of self-reported condom use.

PIP: The objective was to replicate a Zaire study with a cohort of 273 Cameroonian sex workers to determine whether gonorrhea increased the risk of HIV acquisition, and whether the crude association between gonorrheal infection and HIV acquisition was weakened when unprotected coitus was more carefully controlled. 303 Cameroonian female sex workers were enrolled in Yaounde between 1989 and 1990. Eligibility criteria included age 18 years or older; no pregnancy during the previous 42 days; no history of adverse reaction to a spermicidal product; negative enzyme-linked immunosorbent assay (ELISA) HIV-antibody test; and monthly follow-up visits for 1 year. Women were asked to use condoms and suppositories containing N-9 at every sexual activity and coital logs were reviewed monthly. 17 women were identified as cases and 68 as controls. Unlike in Zaire, where 8% of the cases and controls reported never using condoms, none of the women reported unprotected coitus more than 50% of the time in Cameroon. Almost 30% of both cases and controls in Cameroon had evidence of genital ulcers, compared with less than 5% of the cases and controls in Zaire. The crude OR of HIV infection among those who had gonorrhea during the exposure period was much higher in Zaire than in Cameroon (6.3 vs. 2.2). In both the Zaire and Cameroon data the crude OR were reduced (6.3 to 4.8 and 2.2 to 1.7, respectively) by controlling for risk factors of young age, number of partners per week, trichomoniasis, genital ulcers, and a dichotomous variable indicating irregular or no condom use. In the 1st alternative model, the OR for gonorrhea was 2.0 when the dichotomous measure of unprotected coitus (i.e., more than 25% of coital acts were unprotected) was used. When a more precise continuous estimate of level of unprotected coitus was used (i.e., the percentage of acts where neither condoms nor N-9 was used) the OR for gonorrhea was reduced to 1.4.