Incident and prevalent herpes simplex virus type 2 infection increases risk of HIV acquisition among women in Uganda and Zimbabwe : AIDS

Secondary Logo

Journal Logo

BASIC SCIENCE

Incident and prevalent herpes simplex virus type 2 infection increases risk of HIV acquisition among women in Uganda and Zimbabwe

Brown, Joelle Ma; Wald, Annad; Hubbard, Alanc; Rungruengthanakit, Kittipongg; Chipato, Tsungaih; Rugpao, Sungwalg; Mmiro, Francisf; Celentano, David Di; Salata, Robert Sj; Morrison, Charles Sk; Richardson, Barbra Ae; Padian, Nancy Sb

Author Information
doi: 10.1097/QAD.0b013e3282004929
  • Free

Abstract

Introduction

Herpes simplex virus type 2 (HSV-2) is a common sexually transmitted infection, and the most common cause of genital ulcers worldwide [1–6]. HSV-2 infection is highly prevalent in most regions experiencing severe HIV epidemics [7], and the biological features of HSV-2 infection suggest that it may facilitate HIV acquisition. Mucosal herpetic reactivation is associated with an influx of CD4-bearing lymphocytes [8–10], resulting in a larger number of HIV target cells in the genital tracts of persons with HSV-2 infection.

In many developing countries, the public health importance of HSV-2 infection relates to its potential role in facilitating HIV acquisition. Numerous epidemiological investigations have demonstrated the association of genital ulcers in general, and HSV-2 infection in particular, with HIV [11,12]. Eight prospective studies among men and 10 prospective studies among women have evaluated whether HSV-2 infection modifies the risk for HIV acquisition. Overall, the prospective studies among men have shown an increased risk of acquiring HIV in HSV-2-seropositive heterosexual and homosexual men [meta-analysis odds ratio (OR), 2.1; 95% confidence interval (CI), 1.3–3.2] [11]. In contrast, the prospective studies among women have had mixed results; while three studies have found an increased risk for HIV among HSV-2 seropositive women, seven studies have found no statistically significant increased risk [12].

Several possible explanations exist for this sex difference. First, a weaker effect of HSV-2 in women might be expected because of the underlying transmission dynamics of HIV; the transmission probability of HIV from men to women is higher than from women to men [13,14] and, therefore, the baseline probability of infection in women exposed to HIV-positive partners is probably several times higher than in men. Consequently, even if the absolute effects of HSV-2 were similar, lower relative risks would be observed [15]. Second, most of the analyses that have evaluated the risk conferred by HSV-2 infection among women have had relatively small sample sizes and incomplete control of confounding, thereby limiting their ability to detect a significant effect of HSV-2 infection.

Given the strong biological plausibility and the lack of consistent epidemiological data among women, the present study measured incident HIV infection among HSV-2-seropositive and HSV-2-seronegative subjects participating in a large cohort study evaluating the effect of hormonal contraceptive use on HIV acquisition [the Hormonal Contraception and HIV (HC-HIV) Study] [16]. The primary aim of this analysis is to measure the effects of prevalent and incident HSV-2 infection on HIV acquisition.

Methods

The research was approved by the institutional review boards of the Ugandan Ministry of Health, the Medical Research Council of Zimbabwe, the University of California, San Francisco, Case Western Reserve University, Chiang Mai University, and Johns Hopkins University. All women provided written informed consent prior to study participation.

Cohort and study procedures

The HC-HIV Study cohort and procedures have been described in detail elsewhere [16]. In brief, between November 1999 and January 2004, we recruited and followed a nonprobability quota sample of consecutive women seeking reproductive and general healthcare services from three sites in Kampala, Uganda, and four sites in Harare and Chitungwiza, Zimbabwe. Participants were sexually active, aged 18–35 years, HIV negative at screening, not pregnant, and using combined oral contraceptives (low-dose, monophasic pills containing 30 μg ethinylestradiol and 150 μg levonorgestrel), 150 mg depot-medroxyprogesterone acetate (DMPA) administered every 12 weeks, or were not using hormonal contraceptives (condoms, withdrawal, traditional methods, sterilization, no method) for at least 3 months. Each site enrolled roughly equal numbers of women using combined oral contraceptives, DMPA and nonhormonal methods.

At enrollment, women completed a standardized interview to assess sociodemographic characteristics (age, years of education, marital status), lifetime history of sexual partners and sexually transmitted infections (STI), lifetime reproductive history (parity, contraceptive use), sexual behavior over the past 3 months (frequency of sexual intercourse, number of sexual partners, condom use), and characteristics of their current sexual partner (age, circumcision status, presence of genital ulcers). We provided contraceptive counseling, HIV/STI risk reduction counseling, condom use counselling, provided condoms and hormonal contraception. Study clinicians conducted a standardized physical (including pelvic) examination, noting abnormalities in the external genitalia, and the vaginal and cervical epithelium. Blood was collected for HSV-2, HIV, and syphilis testing. Urine was collected for pregnancy testing. Cervicovaginal specimens were collected to test for Chlamydia trachomatis, Neisseria gonorrhoeae, Trichomonas vaginalis, Candida albicans and bacterial vaginosis. Participants were treated onsite for vaginal infections, while those subsequently diagnosed with other treatable STI were recalled for treatment. Women who tested positive for HIV were offered ongoing counseling by study staff and were referred for medical care and emotional support for their HIV infection.

Study follow-up was limited to women who were HIV negative at baseline. Follow-up visits were conducted every 12 weeks for 15–24 months, depending on enrollment date. Follow-up procedures were similar to those at enrollment.

Laboratory methods

HSV-2 status was determined by a type-specific serological IgG antibody enzyme-linked immunosorbent assay (ELISA; Focus Technologies, Cypress, California, USA). For quality control purposes, an external research laboratory repeated the serological testing on a randomly selected 10% of enrollment specimens and compared these results with the University of Washington Western blot as the ‘gold standard’. The methods and results of the quality control testing have been published elsewhere [17]. In order to optimize the specificity of the ELISA in Uganda, an index cutoff value of 3.4 was used to determine HSV-2 seropositivity [18]. All of the HSV-2 seroconversions were confirmed by ELISA at an external laboratory. HSV-2-positive serostatus was defined as the date of the first positive HSV-2 ELISA result.

Because the frequency of HSV-2 reactivations tends to decrease over time [19,20] we hypothesized that risk for HIV due to HSV-2 infection would be greater among women with more recent HSV-2 infection (i.e., incident HSV-2 infection during follow-up) than among those with established HSV-2 infection (i.e., prevalent HSV-2 infection at enrollment). Consequently, the analyses used a three-part time-varying variable for HSV-2 infection that indicated, at each study visit, whether a woman was HSV-2 negative or positive, and if positive, whether she acquired HSV-2 during the study (i.e., incident HSV-2 infection) or tested HSV-2 positive at enrollment (i.e., prevalent HSV-2 infection). Women who acquired HSV-2 infection during the study follow-up period were classified as being HSV-2 negative at each visit prior to HSV-2 seroconversion and classified as having incident HSV-2 at the seroconversion visit and at each subsequent visit.

At enrollment, all participants were determined to be HIV negative by ELISA. At each quarterly follow-up visit, HIV testing was done using ELISA, with positive results confirmed using rapid testing, then with Western blot or polymerase chain reaction (PCR) testing; HIV DNA PCR results were the final arbiter of infection status. Incident HIV infections were confirmed by a subsequent blood draw using an ELISA or rapid test. For confirmed HIV infections, HIV DNA PCR was performed serially on previous visit specimens. HIV incidence was defined as the date of the first positive PCR result, and participants were censored at the time of HIV seroconversion.

Statistical analyses

Logistic regression, adjusted for repeated observations [21], was used to explore bivariable associations between baseline and time-varying characteristics and HIV acquisition. Median and interquartile ranges were compared for continuous variables, and the proportion of person-time for categorical variables was compared between participants who acquired HIV and those who did not. Bivariable associations with HSV-2 serostatus (HSV-2 prevalent, HSV-2 incident, and HSV-2 negative) were computed in a similar fashion using multinomial logistic regression.

In the analysis evaluating risk for HIV due to HSV-2 infection, time was treated as discrete blocks of 3-month intervals. Women contributed a data point for each 3-month interval of time until they tested positive for HIV. The outcome of interest was time to HIV seroconversion. A marginal structural discrete survival model (MSDSM) was used to evaluate prevalent and incident HSV-2 infections as risk factors for acquiring HIV, adjusting for confounding. This approach was chosen because, in the presence of time-dependent confounding [which occurs when time-varying variables (such as having an STI) are related to past HSV-2 status as well as future HSV-2 status and HIV status], unbiased estimates of the parameter of interest (e.g., risk of HIV acquisition) cannot be obtained from standard analytical approaches (e.g., time-varying Cox proportional hazard models) [22].

The MSDSM, which is represented by the following equation,

estimates the hazard of HIV infection, λa(t), when the entire population has HSV-2 status a (either incident, HSVinc, or prevalent, HSVprev). Under certain assumptions (e.g., no unmeasured confounding) [23], the MSDSM can be interpreted as estimating the change in the underlying rate of HIV in a population as the distribution of HSV-2 infection is changed. Thus, β1 and β2 approximate the relative hazards of HIV infection. To correct for confounding, the regression coefficients (β1 and β2) were estimated using the stabilized ‘inverse probability of treatment weighted estimator’ [24]. Technically, the MSDSM was estimated by fitting a weighted logistic regression treating each person interval as an observation [25]. The weights were generated using multinomial logistic regression of HSV-2 status versus potential confounders, which included time-varying and time-invariant variables that were considered a priori to be potential confounders (e.g., age, having a new sex partner, frequency of unprotected sex, hormonal contraceptive use, having a genital ulcer, and testing positive for other STI), and covariates that were associated (P < 0.05) with HIV in bivariable analyses [26]. Including the weights in the MSDSM has the effect of ‘correcting’ for confounding. Generalized estimating equations were used to derive confidence intervals [21].

Because it is not possible to determine the relative timing of HSV-2 and HIV infections that occur during the same interval, women who acquired both infections during the same interval were excluded from all analyses (three women in Uganda and four women in Zimbabwe).

Estimates from the MSDSM were used to approximate the population attributable risk percent (PAR%) of HIV seroconversions due to prevalent and incident HSV-2 infection [27]. Specifically, the PAR% was estimated by subtracting the adjusted probability of acquiring HIV given that women were HSV-2 negative from the overall probability of acquiring HIV in this cohort. This difference was divided by the overall probability of acquiring HIV in this cohort, and then multiplied by 100. The PAR% can be interpreted as the fraction of HIV seroconversions that could have been avoided if all participants had been HSV-2 negative, controlling for confounding.

All analyses were conducted using Stata version 9.0 software (Stata Corporation, College Station, Texas, USA).

Results

A total of 8346 women were screened in order to enroll 4531 women: 2235 in Uganda and 2296 in Zimbabwe. Participant retention at 24 months was 92%: 96% in Uganda and 88% in Zimbabwe. Mean follow-up was 21.9 months; median time between study visits was 81 days. At enrollment, the median age was 25 years; median education was 10 years; median number of lifetime pregnancies was two; and most participants (83%) lived with their primary sexual partner. Few participants reported having more than one sex partner (4%) in the 3 months prior to enrollment, and less than half (42%) reported any condom use in the month prior to enrollment. The characteristics of participants measured at enrollment and during follow-up are shown in Tables 1 and 2 for women from Uganda and Zimbabwe, respectively.

T1-3
Table 1:
Bivariable associations with herpes simplex 2 infection and HIV infection in Uganda: characteristics of participants measured at enrollment and during follow-up.
T2-3
Table 2:
Bivariable associations with herpes simplex 2 infection and HIV infection in Zimbabwe: characteristics of participants measured at enrollment and during follow-up.

Prevalence of herpes simplex virus 2 infection

Approximately half of the study participants tested HSV-2 seropositive at enrollment: 51.5% in Uganda and 53.2% in Zimbabwe. At enrollment, 24% of Ugandan women and 8% of Zimbabwean women reported having ever had a genital ulcer. Of the women who tested HSV-2 positive at enrollment, 27% in Uganda and 13% in Zimbabwe reported having ever had a genital ulcer.

Incidence of herpes simplex virus 2 infection

Overall, 324 HSV-2 seroconversions were detected for an incidence rate of 9.6/100 person-years in Uganda (179 seroconversions) and 8.8/100 person-years in Zimbabwe (145 seroconversions). In Uganda, 530 women (24%) reported having at least one genital ulcer during the study follow-up period. Of these 530 women, 302 (57%) had prevalent HSV-2; 18 (3%) had incident HSV-2; and 210 (40%) were HSV-2 negative. In Zimbabwe, 188 women (12%) reported having at least one genital ulcer during follow-up. Of these 188 women, 153 (81%) had prevalent HSV-2, 10 (5%) had incident HSV-2, and 25 (13%) were HSV-2 negative.

Characteristics associated with prevalent and incident herpes simplex virus 2 infection

In Uganda, characteristics at enrollment associated with having prevalent HSV-2 infection included being older, less educated, married, and having a history of genital ulcers (Table 1). During follow-up, Ugandan women with prevalent HSV-2 were more likely to use hormonal contraception, have unprotected sex, have an uncircumcised partner, engage in commercial sex, report having a genital ulcer, test positive for N. gonorrhoeae, have pelvic inflammatory disease, and were less likely to be pregnant. Ugandan women who acquired HSV-2 during follow-up were less educated, more likely to be in a polygynous marriage, and more likely to report during the interval prior to acquiring HSV-2 to have more sexual partners or an older sex partner, and were more likely during the visit at which they seroconverted to HSV-2 also to be pregnant, and to test positive for N. gonorrhoeae and candidiasis.

In Zimbabwe, characteristics at enrollment associated with having prevalent HSV-2 infection included being older, less educated, having more lifetime sexual partners, and reporting a history of genital ulcers (Table 2). During follow-up, Zimbabwean women with prevalent HSV-2 infection were more likely to have a new partner, an older sexual partner, a partner with a genital ulcer, engage in commercial sex, report having a genital ulcer, and test positive for T. vaginalis, C. trachomatis, N. gonorrhoeae, and bacterial vaginosis. Zimbabwean women who acquired HSV-2 during follow-up tended to be unmarried and were more likely to report, during the interval prior to acquiring HSV-2, engaging in commercial sex or having a genital ulcer, and were more likely during the visit at which they seroconverted to HSV-2 also to test positive for T. vaginalis, C. trachomatis, N. gonorrhoeae, and bacterial vaginosis, and to have a genital ulcer by clinical examination.

Incidence of HIV infection

There were a total of 211 HIV seroconversions, resulting in an HIV incidence rate of 1.6/100 person-years in Uganda (60 HIV seroconversions) and 4.1/100 person-years in Zimbabwe (151 HIV seroconversions). In both countries, HIV incidence was higher among women with HSV-2 infection compared with women without HSV-2. In Uganda, HIV incidence rates among women without HSV-2, with prevalent HSV-2 infection, and with incident HSV-2 infection, were 0.7, 2.0, and 2.9/100 person-years, respectively (Fig. 1). In Zimbabwe, HIV incidence rates among women without HSV-2, with prevalent HSV-2 infection, and with incident HSV-2 infection were 1.2, 5.8, and 9.7/100 person-years, respectively (Fig. 1). Overall, approximately 84% of all HIV seroconversions (178/211) occurred among women with prior HSV-2 infection.

F1-3
Fig. 1:
Crude incidence rates with 95% confidence interval for HIV (excluding three women in Uganda and four women in Zimbabwe who seroconverted to HIV and herpes simplex 2 (HSV-2) during the same interval) stratified by HSV-2 sero status.

Characteristics associated with HIV acquisition

In bivariable analyses, Ugandan women who acquired HIV were younger and more likely to be HSV-2 positive. During the interval prior to acquiring HIV, Ugandan women were more likely to report having had a new sexual partner, and during the visit at which they tested positive for HIV, were more likely to test positive for N. gonorrhoeae, bacterial vaginosis, and have a genital ulcer by clinical examination (Table 1). Zimbabwean women who acquired HIV (Table 2) were less likely to be in a monogamous marriage and more likely to be HSV-2 positive. During the interval prior to acquiring HIV, Zimbabwean women were more likely to report commercial sex, and during the visit at which they tested positive for HIV, they were more likely to test positive for T. vaginalis, C. trachomatis, N. gonorrhoeae, bacterial vaginosis, and candidiasis.

The relationship between prior herpes simplex virus 2 infection and HIV acquisition

In crude and multivariable models, women with prior HSV-2 infection were at significantly greater risk for HIV acquisition (Table 3). The hazard rate ratio (HR) for HIV among Ugandan women with prior HSV-2 infection, compared with women without HSV-2 infection, was 3.2 (95% CI, 1.7–6.3), after controlling for confounding. The HR for HIV among Zimbabwean women with prior HSV-2 infection, compared with women without HSV-2 infection, was 4.6 (95% CI, 2.8–7.6), after adjusting for confounding. The PAR% for HIV associated with prevalent and incident HSV-2 infection was 42% in Uganda and 65% in Zimbabwe.

T3-3
Table 3:
Hazard ratios for HIV infection associated with prevalent and incident herpes simplex virus 2 infection: crude (unweighted) and controlled for confounding (weighted).

The relationship between prevalent and incident herpes simplex virus 2 infection and HIV acquisition

The effect of prevalent HSV-2 infection and incident HSV-2 infection on HIV acquisition were also evaluated separately. In both crude and multivariable analyses, both prevalent HSV-2 infection and incident HSV-2 infection were significantly associated with increased risk for HIV acquisition. The HR values for HIV among Ugandan women with prevalent HSV-2 infection and incident HSV-2 infection, compared with women without HSV-2, were 2.8 (95% CI, 1.5–5.3) and 4.6 (95% CI, 1.6–13.1), respectively, after adjusting for confounding. The HR values for HIV among Zimbabwean women with prevalent HSV-2 infection and incident HSV-2 infection, compared with women without HSV-2, were 4.4 (95% CI, 2.7–7.2) and 8.6 (95% CI, 4.3–17.1), respectively, after adjusting for confounding.

Discussion

Our study adds to the growing body of literature examining the effect of HSV-2 infection on HIV acquisition among women. This is the first longitudinal study to provide evidence that both prevalent and incident HSV-2 infections are significant risk factors for HIV acquisition among women from the general population. The PAR% estimates are consistent with the hypothesis that HSV-2 is responsible for a large proportion of new HIV infections in Africa; we estimate that 42–65% of new HIV infections could be averted if all women remained HSV-2 negative.

The roughly 1.5-fold greater risk for HIV among those recently seroconverting to HSV-2 compared with those with seroprevalent HSV-2 infection at enrollment is biologically plausible. HSV-2 reactivation, which is associated with an influx of HIV target cells to the genital tracts of persons with HSV-2, is most severe close to the time of infection [19,20]. However, we cannot exclude the possibility that new acquisition of HSV-2 infection was a marker for having a new partner infected with both HSV-2 and HIV, as HIV-infected persons reactivate HSV-2 more frequently than healthy persons and so are most likely more infectious for HSV-2. Similar findings were observed in a previous longitudinal study conducted among men in Zimbabwe [28], and among women in Uganda [15].

Although the risk for HIV was 1.5-fold higher among women with incident compared with prevalent HSV-2 infection, the PAR%, which suggests between 42–65% of new HIV infections in this cohort could have been avoided if all women had been HSV-2 negative, is largely driven by prevalent rather than incident. HSV-2 infection. This is because of the high population prevalence of HSV-2 infection. These findings underscore the important role of controlling both prevalent and incident HSV-2 infection as a means of preventing transmission of HIV, and it may be used to inform discussions about the role of HSV-2 testing in populations experiencing an HIV epidemic.

Strengths of this analysis include the prospective design, which allows determination of the timing of HSV-2 infection and HIV infection, the large sample size, and the use of marginal structural models, which allows control for time-dependent confounding by measured covariates. A weakness of the study is that we do not have the male partners' serostatus for HSV-2 and HIV, which could have resulted in potentially important unmeasured confounding. To minimize the possibility that any association found between HSV-2 and HIV occurred because women acquired both infections at the same time, participants who acquired both HSV-2 and HIV during the same interval were excluded from analyses.

Three options are available for controlling HSV-2 spread that may reduce HIV transmission. First, promotion of condom use is indicated; condom use prevents HIV transmission [29,30] and may limit HSV-2 transmission [31–33]. Second, suppressive treatment of HSV-2 infections reduces the frequency, duration, and severity of symptomatic and asymptomatic herpetic reactivations, and reduces transmission of HSV-2 infections to uninfected partners [34]. Antiviral suppression of HSV-2 infections, therefore, has the potential to reduce the risk of HIV indirectly by preventing transmission of HSV-2 and may reduce the risk of HIV acquisition by reducing HSV-2 reactivations. Third, there is preliminary evidence that patients who disclose their HSV-2 status to their partners are less likely to transmit the virus [35], suggesting that serological testing and counseling may result in decreased risk of HSV-2 transmission to sex partners.

Additional methods for HSV-2 control should be explored, including therapeutic and prophylactic HSV-2 vaccines [36,37]. Male circumcision, which has recently been shown to be effective in prevention of HIV acquisition among men [38], may also prevent the acquisition of other STI in men, including HSV-2. The development of a topical microbicide is another option currently being explored [39–48]. The ideal microbicide would prevent both transmission and acquisition of HIV and HSV-2 and, therefore, could play an important role in curbing the spread of HIV, as well as the spread of HSV-2 infection.

Acknowledgements

We would like to thank the study participants in Uganda and Zimbabwe for graciously volunteering their time and information. We would also like to acknowledge the vital contributions of our laboratory teams in the United States, Uganda, and Zimbabwe, and in particular the competence and generosity shown by Jeanne Moncada, Henry Bakka, Kittipong Rungruengthanakit and Marshall Munjoma. We would also like to thank Dr Starley Shade and Dr Ward Cates for their insightful reviews of this manuscript. Finally, this study would not have been possible without the tremendous dedication and talent of the HC-HIV study team.

We declare that we have no conflict of interest. Dr Wald has received grant support from National Institutes of Health, GlaxoSmithKline, Antigenics, 3M, Roche, and Vical. She is a consultant for Novartis, Powdermed, and Medigene and a speaker for Merck.

The data included in this manuscript were generated from a multisite international collaboration. Each of the authors has substantially contributed to the study's conception, design, or performance, and has contributed to the manuscript.

Sponsorship: This study was funded by the National Institute of Child Health and Human Development (NICHD), National Institutes of Health, US Department of Health and Human Services through a contract with Family Health International (N01-HD-0-3310). In addition, Dr Wald was supported by NIAID AI-30731 and AI-071113. NICHD reviewed the study design as well as this manuscript. NICHD was not involved in the collection, analysis, and interpretation of the data.

References

1. Behets FM, Brathwaite AR, Hylton-Kong T, Chen CY, Hoffman I, Weiss JB, et al. Genital ulcers: etiology, clinical diagnosis, and associated human immunodeficiency virus infection in Kingston, Jamaica. Clin Infect Dis 1999; 28:1086–1090.
2. Beyrer C, Jitwatcharanan K, Natpratan C, Kaewvichit R, Nelson KE, Chen CY, et al. Molecular methods for the diagnosis of genital ulcer disease in a sexually transmitted disease clinic population in northern Thailand: predominance of herpes simplex virus infection. J Infect Dis 1998; 178:243–246.
3. Chen CY, Ballard RC, Beck-Sague CM, Dangor Y, Radebe F, Schmid S, et al. Human immunodeficiency virus infection and genital ulcer disease in South Africa: the herpetic connection. Sex Transm Dis 2000; 27:21–29.
4. Fleming DT, McQuillan GM, Johnson RE, Nahmias AJ, Aral SO, Lee FK, et al. Herpes simplex virus type 2 in the United States, 1976 to 1994. N Engl J Med 1997; 337:1105–1111.
5. Limpakarnjanarat K, Mastro TD, Saisorn S, Uthaivoravit W, Kaewkungwal J, Korattana S, et al. HIV-1 and other sexually transmitted infections in a cohort of female sex workers in Chiang Rai, Thailand. Sex Transm Infect 1999; 75:30–35.
6. Morse SA, Trees DL, Htun Y, Radebe F, Orle KA, Dangor Y, et al. Comparison of clinical diagnosis and standard laboratory and molecular methods for the diagnosis of genital ulcer disease in Lesotho: association with human immunodeficiency virus infection. J Infect Dis 1997; 175:583–589.
7. WHO/UN AIDS. Herpes Simplex Virus Type 2: Programmatic and Research Priorities in Developing Countries. Geneva: World health Organization; 2001.
8. Cunningham AL, Nelson PA, Fathman CG, Merigan TC. Interferon gamma production by herpes simplex virus antigen-specific T cell clones from patients with recurrent herpes labialis. J Gen Virol 1985; 66(Pt 2):249–258.
9. Koelle DM, Corey L, Burke RL, Eisenberg RJ, Cohen GH, Pichyangkura R, et al. Antigenic specificities of human CD4+ T-cell clones recovered from recurrent genital herpes simplex virus type 2 lesions. J Virol 1994; 68:2803–2810.
10. Corey L, Wald A. Sexually Transmitted Diseases. 3rd edn. New York: McGraw-Hill; 1999.
11. Wald A, Link K. Risk of human immunodeficiency virus infection in herpes simplex virus type 2-seropositive persons: a meta-analysis. J Infect Dis 2002; 185:45–52.
12. Freeman EE, Weiss HA, Glynn JR, Cross PL, Whitworth JA, Hayes RJ. Herpes simplex virus 2 infection increases HIV acquisition in men and women: systematic review and meta-analysis of longitudinal studies. AIDS 2006; 20:73–83.
13. Nicolosi A, Musicco M, Saracco A, Lazzarin A. Risk factors for woman-to-man sexual transmission of the human immunodeficiency virus. Italian Study Group on HIV Heterosexual Transmission. J Acquir Immune Defic Syndr 1994; 7:296–300.
14. Carpenter LM, Kamali A, Ruberantwari A, Malamba SS, Whitworth JA. Rates of HIV-1 transmission within marriage in rural Uganda in relation to the HIV sero-status of the partners. AIDS 1999; 13:1083–1089.
15. del Mar Pujades Rodriguez M, Obasi A, Mosha F, Todd J, Brown D, Changalucha J, et al. Herpes simplex virus type 2 infection increases HIV incidence: a prospective study in rural Tanzania. AIDS 2002; 16:451–462.
16. Morrison C, Richardon B, Mmiro F, Chipato T, Celentano D, Luoto J, et al. Hormonal contraception and risk of acquiring HIV infection. AIDS 2007; 21:85–95.
17. Hogrefe W, Su X, Song J, Ashley R, Kong L. Detection of herpes simplex virus type 2-specific immunoglobulin G antibodies in African sera by using recombinant gG2, Western blotting, and gG2 inhibition. J Clin Microbiol 2002; 40:3635–3640.
18. Laeyendecker O, Henson C, Gray RH, Nguyen RH, Horne BJ, Wawer MJ, et al. Performance of a commercial, type-specific enzyme-linked immunosorbent assay for detection of herpes simplex virus type 2-specific antibodies in Ugandans. J Clin Microbiol 2004; 42:1794–1796.
19. Koelle DM, Benedetti J, Langenberg A, Corey L. Asymptomatic reactivation of herpes simplex virus in women after the first episode of genital herpes. Ann Intern Med 1992; 116:433–437.
20. Benedetti JK, Zeh J, Corey L. Clinical reactivation of genital herpes simplex virus infection decreases in frequency over time. Ann Intern Med 1999; 131:14–20.
21. Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika 1986; 73:13–22.
22. Robins JM, Hernan MA, Brumback B. Marginal structural models and causal inference in epidemiology. Epidemiology 2000; 11:550–560.
23. van der Laan M, Robins J. Unified Methods for Censored Longitudinal Data and Causality. New York: Springer; 2003.
24. Robins JM. Association, causation, and marginal structural models. Synthese 1999; 121:151–179.
25. Hernan MA, Brumback B, Robins JM. Marginal structural models to estimate the causal effect of zidovudine on the survival of HIV-positive men. Epidemiology 2000; 11:561–570.
26. Fisher R. Statistical Methods for Research Workers. 13th edn. Edinburgh: Oliver & Boyd; 1932.
27. Hubbard A, van der Laan M. UC Berkeley Division of Biostatistics Working Paper Series 151: Population Intervention Models in Causal Inference. Berkeley, CA: University of California Press; 2005.
28. McFarland W, Gwanzura L, Bassett MT, Machekano R, Latif AS, Ley C, et al. Prevalence and incidence of herpes simplex virus type 2 infection among male Zimbabwean factory workers. J Infect Dis 1999; 180:1459–1465.
29. de Vincenzi I. A longitudinal study of human immunodeficiency virus transmission by heterosexual partners. European Study Group on Heterosexual Transmission of HIV. N Engl J Med 1994; 331:341–346.
30. Cayley WE Jr. Effectiveness of condoms in reducing heterosexual transmission of HIV. Am Fam Physician 2004; 70:1268–1269.
31. Wald A, Langenberg AG, Krantz E, Douglas JM Jr, Handsfield HH, DiCarlo RP, et al. The relationship between condom use and herpes simplex virus acquisition. Ann Intern Med 2005; 143:707–713.
32. Wald A, Langenberg AG, Link K, Izu AE, Ashley R, Warren T, et al. Effect of condoms on reducing the transmission of herpes simplex virus type 2 from men to women. JAMA 2001; 285:3100–3106.
33. Gottlieb SL, Douglas JM Jr, Foster M, Schmid DS, Newman DR, Baron AE, et al. Incidence of herpes simplex virus type 2 infection in 5 sexually transmitted disease (STD) clinics and the effect of HIV/STD risk-reduction counseling. J Infect Dis 2004; 190:1059–1067.
34. Corey L, Ashley R. Prevention of herpes simplex virus type 2 transmission with antiviral therapy. Herpes 2004; 11:170A–174A.
35. Wald A, Krantz E, Selke S, Lairson E, Morrow RA, Zeh J. Knowledge of partners' genital herpes protects against herpes simplex virus type 2 acquisition. J Infect Dis 2006; 194:42–52.
36. Stanberry LR. Clinical trials of prophylactic and therapeutic herpes simplex virus vaccines. Herpes 2004; 11(Suppl 3):161A–169A.
37. Garnett GP, Dubin G, Slaoui M, Darcis T. The potential epidemiological impact of a genital herpes vaccine for women. Sex Transm Infect 2004; 80:24–29.
38. Auvert B, Taljaard D, Lagarde E, Sobngwi-Tambekou J, Sitta R, Puren A. Randomized, controlled intervention trial of male circumcision for reduction of HIV infection risk: the ANRS 1265 Trial. PLoS Med 2005; 2911:e298.
39. Palliser D, Chowdhury D, Wang QY, Lee SJ, Bronson RT, Knipe DM, et al. An siRNA-based microbicide protects mice from lethal herpes simplex virus 2 infection. Nature 2006; 439:89–94.
40. Johnson DC. Silencing herpes simplex virus with a vaginal microbicide. N Engl J Med 2006; 354:970–971.
41. Keller MJ, Zerhouni-Layachi B, Cheshenko N, John M, Hogarty K, Kasowitz A, et al. PRO 2000 gel inhibits HIV and herpes simplex virus infection following vaginal application: a double-blind placebo-controlled trial. J Infect Dis 2006; 193:27–35.
42. Anderson RA, Feathergill KA, Diao XH, Cooper MD, Kirkpatrick R, Herold BC, et al. Preclinical evaluation of sodium cellulose sulfate (Ushercell) as a contraceptive antimicrobial agent. J Androl 2002; 23:426–438.
43. Carlucci MJ, Scolaro LA, Noseda MD, Cerezo AS, Damonte EB. Protective effect of a natural carrageenan on genital herpes simplex virus infection in mice. Antiviral Res 2004; 64:137–141.
44. Bernstein DI, Stanberry LR, Sacks S, Ayisi NK, Gong YH, Ireland J, et al. Evaluations of unformulated and formulated dendrimer-based microbicide candidates in mouse and guinea pig models of genital herpes. Antimicrob Agents Chemother 2003; 47:3784–3788.
45. Zeitlin L, Hoen TE, Achilles SL, Hegarty TA, Jerse AE, Kreider JW, et al. Tests of Buffergel for contraception and prevention of sexually transmitted diseases in animal models. Sex Transm Dis 2001; 28:417–423.
46. Maguire RA, Bergman N, Phillips DM. Comparison of microbicides for efficacy in protecting mice against vaginal challenge with herpes simplex virus type 2, cytotoxicity, antibacterial properties, and sperm immobilization. Sex Transm Dis 2001; 28:259–265.
47. Piret J, Lamontagne J, Bestman-Smith J, Roy S, Gourde P, Desormeaux A, et al. In vitro and in vivo evaluations of sodium lauryl sulfate and dextran sulfate as microbicides against herpes simplex and human immunodeficiency viruses. J Clin Microbiol 2000; 38:110–119.
48. Zaneveld LJ, Waller DP, Anderson RA, Chany C 2nd, Rencher WF, Feathergill K, et al. Efficacy and safety of a new vaginal contraceptive antimicrobial formulation containing high molecular weight poly(sodium 4-styrenesulfonate). Biol Reprod 2002; 66:886–894.
Keywords:

Africa; HIV; herpes simplex virus 2; marginal structural models; population attributable risk; women

© 2007 Lippincott Williams & Wilkins, Inc.