Risk Factors for Incidence of Sexually Transmitted Infections Among Women in South Africa, Tanzania, and Zambia: Results From HPTN 055 Study : Sexually Transmitted Diseases

Secondary Logo

Journal Logo


Risk Factors for Incidence of Sexually Transmitted Infections Among Women in South Africa, Tanzania, and Zambia: Results From HPTN 055 Study

Kapiga, Saidi MD*; Kelly, Cliff MS; Weiss, Stephen PHD; Daley, Tara MS*; Peterson, Leigh PHD§; Leburg, Corey MHS; Ramjee, Gita PHD

Author Information
Sexually Transmitted Diseases 36(4):p 199-206, April 2009. | DOI: 10.1097/OLQ.0b013e318191ba01
  • Free

THE WORLD HEALTH ORGANIZATION ESTIMATES that about 340 million new cases of the 4 main curable sexually transmitted infections (STIs) (chlamydial infection, gonorrhea, syphilis, and trichomoniasis) occur every year in the world.1 Most infections are in developing countries where STIs are a major public health problem. Women who experience severe adverse reproductive health outcomes, including pelvic inflammatory diseases, infertility, cervical intraepithelial neoplasia, and obstetrics complications,2,3 are disproportionately impacted by STIs. In addition, mounting evidence suggests complex interactions between HIV and other STIs that increase transmission of both infections. Bacterial and viral STIs often increase HIV susceptibility and infectiousness,4 and HIV infection may increase occurrence of other STIs.5,6

Comprehensive STI surveillance is lacking in most sub-Saharan African countries where the true burden of these infections is unknown, yet accurate STI incidence estimates are required in the design of prevention programs for targeted populations. Such information can be obtained from prospective studies involving frequent collection of genital samples and rigorous measurements of potential risk factors. In this report, we present results from such a prospective cohort study conducted by the HIV Prevention Trials Network to assess preparedness of potential phase II/IIB microbicide trial sites in South Africa, Tanzania, and Zambia. As a part of the secondary aims of this study, data were analyzed to determine the incidence and risk factors for chlamydial infection, gonorrhea, syphilis, and trichomoniasis.


Study Population

This cohort study was conducted at 4 sites: Lusaka, Zambia; Moshi, Tanzania; and Durban and Hlabisa, South Africa. Sexually active women were recruited from the general population of Chilenje and Kamwala communities in Lusaka, and from a large family planning clinic and high-risk population of bar and hotel workers in Moshi. South African sites recruited women from the general population in the Durban suburb of Chatsworth and rural communities in Hlabisa. Written informed consent was obtained from all enrolled participants, and the study was approved by the ethics committee of each institution involved in the study.

Study Procedures

During screening, women were interviewed to collect demographic characteristics and other information required for eligibility determination. After successful enrollment interviews, urine samples (pregnancy testing) and blood and genital samples (HIV/STI testing) were taken. Physicians subsequently recorded baseline medical history and performed physical and pelvic examination of all participants. At Lusaka and Hlabisa sites, pelvic exam involved use of a speculum with naked eye examinations, whereas colposcopic examinations were performed at the Moshi and Durban sites.

Women meeting eligibility criteria were interviewed to obtain baseline sexual practice and information about use of contraceptive methods. Participants were followed-up once every month for 12 months. Monthly assessments included a medical/menstrual history and urine pregnancy testing. Quarterly, interviews were conducted to update information about sexual behaviors, samples were collected for HIV/STI testing, and pelvic examinations were performed. At each visit, women received individual HIV/STI risk reduction counseling, including the recommendation to use condoms, which were provided free of cost. Additionally, syndromic treatment for participants presenting signs or symptoms of STIs and/or other genital infections was provided free of charge.

Laboratory Procedures

Saline wet mount of the vaginal smear was prepared and examined microscopically for the presence of clue cells, motile Trichomonas vaginalis, white blood cells, and/or yeast cells. In addition, 10% potassium hydroxide was added to a second smear to test for amine odor and examination of hyphae. Disturbances of vaginal flora and bacterial vaginosis were diagnosed on site on the basis of 3 of 4 clinical criteria as proposed by Amsel et al.7 Air-dried smears of vaginal secretions were Gram stained and examined at the Central Laboratory and scored using the Nugent criteria.8 In Durban and Hlabisa, urine samples were tested for Chlamydia trachomatis and Neisseria gonorrhea using the BD Probe Tec ET assay (Becton Dickinson, MD) according to manufacturers’ instructions. In Moshi and Lusaka, endocervical swabs were collected for detection of C. trachomatis antigen via enzyme-linked immunosorbent assay (ELISA, Murex Biotech), and gonorrhea using culture methods. The BD Probe Tec urine assay has a specificity of more than 97% and a sensitivity of about 80% as compared with culture methods9; chlamydial ELISA has a sensitivity ranging from 65% to 92%, and a specificity of 96% to 100%.10

Active (recent) syphilis was diagnosed at the local laboratories based on positive serum reaction after both a rapid plasma reagin card test and a more specific confirmatory test treponema pallidum haemagglutination assay (TPHA) or microhaemagglutination assay-treponema pallidum (MHA-TP). At screening, HIV status was ascertained using 2 concurrent rapid enzyme immunoassay tests (Abbott Determine and Orasure OraQuick). If the rapid tests were discordant, a Bio-Rad Western blot test was performed to confirm results. Follow-up HIV testing was first performed with Orasure OraQuick rapid enzyme immunoassay test, and positive samples were confirmed with a Bio-Rad Western blot. At the end of the study, all HIV-positive samples detected during follow-up visits and a small proportion of other samples were shipped to the HIV Prevention Trials Network Central Laboratory for confirmatory testing and quality assurance purposes.

Statistical Methods

Data forms from each site were faxed to the Statistical Center for HIV/AIDS Research and Prevention for processing. Incidence rates were calculated as the number of incident events (allowing for multiple events per participant) divided by the total PYAR. Results are presented as number of events/100 PYAR. For the follow-up time, we used discrete time for each quarterly visit (0.25 PYAR). The variance for the CI was calculated based upon methods by Stukel and Glynn.11,12 Because women with STIs and other genital infections at all sites received effective treatment, chlamydial infection, gonorrhea, and trichomoniasis infections detected during the follow-up period were regarded as new infections. For data analysis of syphilis, all confirmed positive serology results from baseline were included, but positive results from follow-up were only included when a positive result occurred after a negative test result.

For each STI, incidence rates were computed for various subgroups defined by socio-demographic characteristics, baseline and follow-up sexual behavior, and other risk factors. To assess associations between risk factors and each STI across sites, we employed the generalized estimating equation method with a logit link and an exchangeable correlation structure. Because of the South African sites performed assays for C. trachomatis and N. gonorrhea with higher levels of sensitivity (BD Probe Tec ET assay) than those used in Tanzania and Zambia (ELISA and culture methods), the associations between risk factors for chlamydial infection and gonorrhea were performed separately for the South Africa sites (combined) and the Tanzania/Zambia sites (combined). Analyses for syphilis and T vaginalis were each performed using data from all 4 sites combined. In the GEE models, factors significant at the 0.20 level in crude models were considered for inclusion in an adjusted model. Backward selection was used to retain only the factors significant at the 0.05 level in the final adjusted model. All statistical analyses were performed using SAS 9.1.3 (SAS Institute, Inc., NC).


Of 2044 women screened for eligibility between June 2003 and October 2004, 958 (47%) were enrolled in the study. The study enrolled 240 women each from Durban and Moshi, and 239 each from Lusaka and Hlabisa. The 12-month retention rates were 97% at Durban, 94% at Hlabisa, 93% at Lusaka, and 86% at Moshi. The mean age of women enrolled in the study was 28.6 years (range, 16–62 years). At the time of enrollment, 46% women were married, whereas 45% were not married but living with a male partner.

The incidence of infection as measured in PYAR was as follows: overall trichomoniasis, 31.9/100 PYAR (95% CI: 27.5–36.3); chlamydial infection in South Africa, 19.5/100 PYAR (95% CI: 14.9–24.1); chlamydial infection in Tanzania and Zambia, 4.9/100 PYAR (95% CI: 2.8–6.9); gonorrhea in South Africa, 16.5/100 PYAR (95% CI: 12.9–20.2); gonorrhea in Tanzania and Zambia, 5.3/100 PYAR (95% CI: 3.2–7.4); and overall syphilis, 7.5/100 PYAR (95% CI: 5.2–9.8). Across all sites the incidence of HIV during the study period was 3.8/100 PYAR (95% CI: 2.6–5.2). The HIV incidence was highest in the 2 South African sites followed by Lusaka, and Moshi, which had the lowest.

In Table 1, we present the incidence of STIs by site and selected baseline socio-demographic and behavioral factors. The highest incidences of chlamydial infection (24.6/100 PYAR) and gonorrhea (20.8/100 PYAR) occurred at the Durban site, where the BD Probe Tec ET assay was used. The Hlabisa site had the highest incidence of trichomoniasis (62.5/100 PYAR), whereas Lusaka had the highest incidence of syphilis (12.1/100 PYAR). Women who had 2 or more sex partners in the past 3 months had higher incidence of STIs with the exception of syphilis and gonorrhea in Moshi/Lusaka. Women who received money or gifts in association with their last incident of vaginal sex had higher rates of chlamydial infection, gonorrhea, and trichomoniasis.

Incidence of Sexually Transmitted Infections (Events/100 Person Years) by Selected Socio-Demographic and Behavioral Factors Among Women in the HPTN 055, Based on 1 Year Follow-Up for Women Enrolled Between May 2003 and October 2004

In Table 2, we illustrate the associations between socio-demographic and behavioral factors and STI incidence. Compared to married women in South Africa, unmarried women with a partner were significantly more likely to have chlamydial infection (OR = 2.3, 95% CI: 1.3–4.1) and gonorrhea (OR = 1.8, 95% CI: 1.0–3.0). Among all sites, unmarried women with a partner were more likely to have trichomoniasis (OR = 2.0, 95% CI: 1.5–2.8). In South Africa, the risk of chlamydial infection (OR = 1.7, 95% CI: 1.0–2.8) and gonorrhea (OR = 2.0, 95% CI: 1.2–3.4) was greater among women who had received higher levels of education (secondary or more). Women who reported earning their own income had a significantly lower risk of trichomoniasis (OR = 0.6, 95% CI: 0.4–0.8). The risk of all STIs except syphilis was increased among women with multiple partners, although this was statistically significant only for chlamydial infection in South Africa. Occurrence of anal sex had significantly increased risk of trichomoniasis (OR = 3.4, 95% CI: 1.2–9.2) and gonorrhea in Tanzania and Zambia (OR = 13.1, 95% CI: 2.0–85.0).

The Associations Between Socio-Demographic and Behavioral Factors and Sexually Transmitted Infections Incidence Among Women in HPTN 055 Study, Based on 1-Year of Follow-Up for Women Enrolled Between May 2003 and October 2004

In Table 3, we illustrate the associations between the incidence of STIs and other genital infections and contraceptive use. Women using oral contraceptive pills and injectable contraceptives showed significantly reduced risk of trichomoniasis. Women with certain STIs or other genital infections were more likely to have an STI during the follow-up period. Gonorrhea and bacterial vaginosis were associated with increased risk of incident trichomoniasis, whereas chlamydial infection in South Africa and syphilis in Tanzania and Zambia were associated with increased risk of incident gonorrhea. The incidence of chlamydial infection was increased among women with gonorrhea in South Africa and among women with candida in Tanzania and Zambia. In general, women with incident HIV infection showed an increased risk of other STIs, although this was only statistically significant in South Africa for chlamydial infection (OR = 7.3, 95% CI: 2.9–18.6) and gonorrhea (OR = 5.5, 95% CI: 1.9–16.3).

The Associations Between Contraceptive Use, Other Genital Infections and Sexually Transmitted Infections Incidence Among Women in HPTN 055 Study, Based on One Year of Follow-Up for Women Enrolled Between May 2003 and October 2004

In Table 4, After adjusting for other predictors in multivariate analysis, we identified a number of independent risk factors for STI. In South Africa, chlamydial infection was influenced by age (older age is protective), number of sex partners, use of injectable contraceptives, gonorrhea, and incident HIV infection. In Tanzania and Zambia, chlamydial infection was influenced by candida and abnormal cervical discharge on exam, and trichomoniasis was influenced by reported occurrence of anal sex in the past 3 months, use of oral contraceptive pills or injectables at time of study (both of which were associated with a lower incidence), gonorrhea, bacterial vaginosis, and abnormal vaginal discharge on exam. In South Africa, gonorrhea was influenced by incident HIV. In Tanzania and Zambia, gonorrhea was influenced by use of Norplant, and syphilis was influenced by age, having a husband or partner who earns an income, number of vaginal sex acts per week (women reporting more sex had a lower incidence), and trichomoniasis.

Results of Multivariate Analysis of Risk Factors for Sexually Transmitted Infections Incidence Among Women in HPTN 055 Study, Based on 1 Year of Follow-Up for Women Enrolled Between May 2003 and October 2004


Results from this study show that STIs are a major problem among women in these populations. Overall, the incidence of most STIs was highest among the 2 South African sites, followed by the Lusaka and Moshi sites, respectively. The higher rates of C. trachomatis and N. gonorrhea observed at the South African sites may be because of the higher sensitivity of the assays used for detection of these infections. A similar pattern was observed for HIV incidence, with the highest rate of infection at the South African and Zambian sites.13 The rate of HIV infection at the South African sites was consistent with results reported recently from a national HIV survey.14 Most women who acquired new STIs were asymptomatic, which is consistent with other studies.15,16 These results highlight the importance of regular STI screening and testing, and underscore the need for treatment as part of standard of care in microbicide efficacy trials and in any population of women with high burden of asymptomatic STIs.

We identified several factors associated with incident STIs in the study population. Measures of high-risk sexual behaviors were associated with increased risk of STIs in this study. Because STIs are known to facilitate HIV transmission and both infections share common modes of transmission, these behaviors seem to be significant contributors to the continued expansion of the HIV epidemic in the study population. As reported in other studies,17–19 sexual behavior measures may be affected by social desirability and misreporting. This could have introduced misclassifications and affected our ability to more accurately assess the associations between STIs and other behavioral variables examined in this analysis.

Women enrolled in the study received HIV/STI counseling, including free condoms, throughout the study period. Despite these efforts, high-risk sexual behaviors were common, and consistent condom use remained low over the course of the study. Thus, we were unsuccessful in reducing high-risk behaviors in this population. The low uptake of condoms in this study highlights the limitations of this method for prevention of HIV among women. Women may not be able to use condoms because they are often unable to negotiate their use with male partners because of fear of abuse or accusations of infidelity.20,21 Additionally, women who are having sex with men in exchange for gifts or money may be reluctant to use condoms if men are willing to pay more for sex without a condom.22 Overall, these findings indicate that development of women-initiated methods for HIV/STI prevention, including microbicidal products, is a critical priority in this population.

We observed strong associations among the different STIs and other genital infections in this study. Women with abnormal vaginal discharge had increased risk of trichomoniasis, and women with abnormal cervical discharge and candidiasis were at increased risk of chlamydial infection in Tanzania and Zambia. The risk of acquiring STIs was also increased among women with these infections during the study. Similarly, incident HIV infections were associated with increased risk of acquisition of chlamydial infection and gonorrhea in South Africa. Among women who acquired both HIV and other STIs, the STI was detected before or at the same visit at which HIV was also detected. These associations provide evidence for the epidemiologic synergy between HIV and other STIs, and are indicative of common behavioral risk factors associated with these infections.

The risk of acquiring trichomoniasis was increased among women with bacterial vaginosis. This is consistent with results from a prospective study conducted in Kenya where abnormal vaginal flora and bacterial vaginosis were associated with an 80% increased risk of trichomoniasis.23 Two other cross-sectional studies have reported significant associations between bacterial vaginosis and trichomoniasis.24,25 Hydrogen peroxide-producing lactobacilli have been shown to directly inhibit STIs, including trichomoniasis.26–28 Thus, reduced local defense against infections in the vaginal microenvironment because of the deficiency of hydrogen peroxide-producing lactobacilli, and presence of anaerobic Gram-negative bacteria, might account for the increased risk of trichomoniasis and other genital infections among women with bacterial vaginosis.

Other predictors of STIs in the study population were use of modern contraceptive methods during the study period. Women using implants had a greater (>4fold) risk of gonorrhea in Tanzania and Zambia, whereas those using oral and injectable contraceptives had marginally reduced risk of trichomoniasis. Injectable contraceptives were also associated with an increased risk of chlamydial infection in South Africa. Hormonal contraceptives have been associated with increased risk of bacterial STIs and reduced risk of trichomoniasis.29 Cervical ectopy is common among women using hormonal contraceptives.30 The increased risk of bacterial STIs among hormonal contraceptive users may be related to cervical ectopy resulting in increased number of receptive cells in the ectocervix for STIs to infect.29 The mechanisms to account for relatively lower risk of trichomoniasis among hormonal contraceptives users are not well established and further research is needed in this area.

The strengths of our study include the prospective cohort design involving women in 3 African countries, along with very high retention rate ranging from 86% to 97% across sites. This allowed for frequent detection of STIs, provision of necessary treatment, and ongoing collection of behavioral risk factors. A limitation of our study is that study subjects were from selected populations and may not be representative of women in the general population. Another limitation is that sexual behaviors may be affected by social desirability bias and misreporting. At all sites, we conducted confidential interviews using trained same-sex interviewers, which might have minimized such issues. Our study may have been further hampered by unexpected STI outbreaks. Although all STIs were treated with highly effective therapy, some may have failed to respond to treatment, and some untreated partners may have contributed to new or other infections. Finally, variance in the sensitivity of diagnostic methods to detect chlamydia and gonorrhea across sites might have resulted in underestimations of infectious incidence at some sites (i.e., Lusaka and Moshi).

In conclusion, the incidence of STIs, including HIV, was high among women in this study. High-risk sexual behaviors were associated with increased risk of STIs, suggesting that effective behavioral interventions are urgently needed. Ongoing counseling and provision of condoms were not effective in promoting safer sexual practices and reducing the incidence of STIs. All of this highlights the need for development of women-initiated methods for HIV/STI prevention, including microbicidal products, and the need to include STI services as part of reproductive health services.


1. World Health Organization. Global prevalence and incidence of selected curable sexually transmitted infections: overview and estimates. Geneva, Switzerland: World Health Organization, 2001.
2. Aral SO, Hawkes S, Biddlecom A, et al. Disproportionate impact of sexually transmitted diseases among women. Emerging Infect Dis 2004; 10:2029–2030.
3. Low N, Broutet N, Adu-Sarkodie Y, et al. Global control of sexually transmitted infections. Lancet 2006; 368:2001–2016.
4. Galvin SR, Cohen MS. The role of sexually transmitted diseases in HIV transmission. Nat Rev Microbiol 2004; 2:33–42.
5. Corey L, Wald A, Celum CL, et al. The effects of herpes simplex virus-2 on HIV-1 acquisition and transmission: A review of two overlapping epidemics. J Acquir Immun Defic Syndr 2004; 35:435–445.
6. Paavonen J, Lehtinen M. Interactions between human papillomavirus and other sexually transmitted agents in the etiology of cervical cancer. Curr Opin Infect Dis 1999; 12:67–71.
7. Amsel R, Totten PA, Spiegel CA, et al. Non-specific vaginitis: Diagnostic criteria and microbial and epidemiological associations. Am J Med 1983; 74:14–22.
8. Nugent RP, Krohn MA, Hillier SL. Reliability of diagnosing bacterial vaginosis is improved by a standardized method of gram stain interpretation. J Clin Microbiol 1991; 29:297–301.
9. Cook RL, Hutchison SL, Ostergaard L, et al. Systematic review: Noninvasive testing for Chlamydia trachomatis and Nisseria gonorrhoeae. Ann Intern Med 2005; 142:914–925.
10. Black CM. Current methods of laboratory diagnosis of Chlamydia trachomatis infections. Clin Microbiol Rev 1997; 10:160–184.
11. Glynn RJ. Ways of measuring rates of recurrent events. BMJ 1996; 312:364–367.
12. Stukel TA, Glynn RJ, Fisher ES, et al. Standardized rates of recurrent events. Stat Med 1994; 13:1781–1791.
13. Ramjee G, Kapiga S, Weiss S, et al; the HPTN 055 study team. The value of site preparedness studies for future implementation of phase 2/IIb/III HIV prevention trials: Experience from the HPTN 055 study. J Acquir Immun Defic Syndr 2008; 47:93–100.
14. Rehle T, Shisana O, Pillay V et al. National HIV incidence measures—new insights into the South African epidemic. S Afr Med J 2007; 97:194–199.
15. Peterman TA, Tian LH, Metcalf CA, et al. High incidence of new sexually transmitted infections in the year following a sexually transmitted infection: A case for rescreening. Ann Intern Med 2006; 145:564–572.
16. Oral SO. Sexually transmitted diseases. Magnitude, determinants, and consequences. Int J STD AIDS 2001; 12:211–215.
17. Buve A, Lagarde E, Carael M, et al. Interpreting sexual behavior data: Validity issues in the multicenter study on factors determining the differential spread of HIV in four African cities. AIDS 2001; 15(suppl 4):S117–S126.
18. Cleland J, Boerma JT, Carael M, et al. Monitoring sexual behavior in general populations: A synthesis of lessons of the past decade. Sex Transm Infect 2004; 80:ii1–ii7.
19. Fishbein M, Pequegnat W. Evaluating AIDS prevention interventions using behavioral and biological outcome measures. Sex Transm Dis 2000; 27:101–110.
20. Kapiga SH, Lwihula GK, Shao JF, et al. Predictors of AIDS knowledge, condom use, and high-risk sexual behavior among women in Dar-es-Salaam, Tanzania. Int J STD AIDS 1995; 6:175–183.
21. Mayaud P, Mabey D. Approaches to the control of sexually transmitted infections in developing countries: Old problems and modern challenges. Sex Transm Infect 2004; 80:174–182.
22. Wellings K, Collumbien M, Slaymaker E, et al. Sexual behavior in context: A global perspective. Lancet 2006; 368:1706–1728.
23. Martin HL, Richardson BA, Nyange PM, et al. Vaginal lactobacilli, microbial flora, and risk of human immunodeficiency virus type 1 and sexually transmitted disease acquisition. J Infect Dis 1999; 180:1863–1868.
24. Moodley P, Connolly C, Sturm AW. Interrelationships among human immunodeficiency virus type 1 infection, bacterial vaginosis, trichomoniasis, and the presence of yeasts. J Infect Dis 2002; 185:69–73.
25. Franklin TL, Monif GR. Trichomonas vaginalis and bacterial vaginosis: Co-existence in vaginal wet mount preparations from pregnant women. J Reprod Med 2000; 45:131–134.
26. McGregor JA, French JI. Bacterial vaginosis in pregnancy. Obstet Gynecol Survey 2000; 55(suppl 1):S1–S19.
27. Hillier SL, Krohn MA, Nugent RP, et al. Characteristics of three vaginal flora patterns assessed by Gram stain among pregnant women. Am J Obstet Gynecol 1992; 166:938–944.
28. Ness RB, Hillier SL, Richter HE, et al. Douching in relation to bacterial vaginosis, lactobacilli, and facultative bacteria in the vagina. Obstet Gynecol 2002; 100:765–772.
29. Carlin EM, Boag FC. Women, contraception, and STDs including HIV. Int J STD AIDS 1995; 6:373–386.
30. Harrison HR, Costin M, Meder JB, et al. Cervical Chlamydia trachomatis infection in university women: Relationship to history, contraception, ectopy, and cervicitis. Am J Obstet Gynecol 1985; 153:244–251.
© Copyright 2009 American Sexually Transmitted Diseases Association