THE DISPARATE PATTERN and scale of HIV epidemics globally has increasingly drawn attention to the interaction between HIV and the “classic” sexually transmitted diseases (STDs), which have a higher prevalence where their treatment is neglected. A consensus has grown that other STDs increase the spread of HIV, a hypothesis first suggested by Piot et al 1 in 1984. Following on from this hypothesis and the early epidemiologic studies, 2–5 several reviews have explored the epidemiologic synergy between STDs and HIV, 6–12 concluding that the presence of either genital ulcer diseases (GUDs) or nonulcerative diseases (NUDs) increase the transmission probability of HIV within a partnership. However, the interaction of the many STDs with HIV is potentially complex, with the possibility of reciprocal influences on susceptibility, infectiousness, and the natural history of infections.
In terms of public health, the most important interaction is the impact of STDs on HIV transmission because this would allow the treatment of STDs to be used as a means of controlling the spread of HIV. However, many interactions are possible, including an increase in STD transmission in the presence of HIV, which could generate positive feedback (referred to as epidemiologic synergy) in the epidemiology of both infections. The different possible interactions between STDs and HIV are outlined in Figure 1. To estimate the magnitude of these influences, data must be collected from different types of clinical and epidemiologic studies. Ideally, these data should be estimated for all the different types of STDs. Both genders should be examined (i.e., both male-to-female and female-to-male transmission, as well as the duration of infectiousness in both women and men).
To base clinical and public health decisions on the best available evidence, standard methods for combining the results of studies have been developed. Initially, such systematic reviews and meta-analyses were used to provide an overall measure of effect for a treatment or exposure. Recently, however, more emphasis has been focused on identification of heterogeneities between studies, careful interpretation of observed differences, and examination of possible publication bias. 13–15 In an effort to identify what is known about the interaction of STDs and HIV in heterosexual couples and what is not known, we conducted a systematic review and meta-analysis, when appropriate, on the basis of currently available studies.
Materials and Methods
A computerized search of the PubMed databases, Medline and PreMedline, of the National Library of Medicine, National Institutes of Health was conducted to identify all studies that had examined the interrelationship of HIV and classic STDs. This search excluded human papillomavirus (HPV) and its diseases, genital warts (condyloma acuminata), and cervical intraepithelial neoplasia because most of this vast literature focuses on the cofactor effect of HIV on HPV–induced cervical intraepithelial neoplasia and not on transmission. Human herpes virus type 8 (the causative agent of Kaposi sarcoma) also was excluded because its route of transmission is unclear.
Search words were selected after pilot searches to check whether pre-identified relevant studies were found in the search. The search term was HIV infections OR HIV AND Sexually Transmitted Diseases[Text word] OR STDs[Text word] OR Sexually Transmitted Diseases, Bacterial OR Herpes genitalis OR Herpes Simplex OR Herpesvirus 1, Human OR Herpesvirus 2, Human OR Syphilis OR Chancre OR Treponema pallidum OR Chancroid OR Haemophilus ducreyi OR Gonorrhea OR Neisseria gonorrhoeae OR Chlamydia infections OR Lymphogranuloma Venereum OR Chlamydia trachomatis OR Trichomonas vaginitis OR Trichomonas vaginalis OR Vaginitis OR Vaginosis, bacterial OR Cervicitis OR Urethritis.
The search was conducted on February 12, 2000, and therefore should include most relevant studies published until January 2000. The search resulted in 5,741 hits. A similar extensive search and review of articles published from 1987 to September 1998 has recently been published, and we initially relied on it to exclude hits before 1998. 12 Therefore, only the abstracts of articles published in 1998, 1999, and 2000 (1,064 reports) were examined in the first search.
In addition, a search was conducted for the first and senior author of already identified relevant studies in the event that these authors had published new studies not captured by the first search. The references of all relevant studies and reviews were examined for additional studies. Finally, the Science Citation Index of the Institute of Scientific Information was used on February 15 to browse the abstracts of all the articles that had cited any of the identified prospective studies.
No efforts were made to identify unpublished studies because no registers of nonexperimental studies exist. However, the investigators of the Mwanza intervention trial were contacted and kindly provided unpublished data. 16 No restriction on language was used, but only English language studies were identified.
Most of the relevant studies have been cross-sectional and case–control investigations. This design has several shortcomings. First, the measurement and classification of the exposure (i.e., the classic STD) may be uncertain because the study is retrospective. Second, the temporal relation is unclear because it is not obvious whether the STD preceded the HIV infection or vice versa. Because of this temporal uncertainty such studies generally examine the summed effects of STDs on susceptibility to HIV and of HIV on susceptibility to STD. Third, because both HIV and STDs are sexually transmitted, it is difficult to control for sexual behavior as a confounder. Many investigators have suggested that higher prevalence of STDs in patients who are HIV-positive than in control subjects who are HIV-negative may be a marker of high-risk behavior rather than evidence of a causal relation. A longitudinal design (i.e., prospective, and some retrospective, cohort studies and nested case–control studies) provides more information on the timing and nature of infection and makes it easier to control for confounding variables, although there inevitably will be some residual confounding.
For the systematic review of effects on susceptibility to HIV or STD infections, only longitudinal studies were included. For the effects on infectiousness, both discordant partnership studies and studies on shedding of HIV or STD pathogens were included because effects on infectiousness cannot be estimated from standard designs. For effects on the duration of the disease, all controlled studies were included.
From each of the identified reports, we extracted data representing study population, study design, length of follow-up period, study size, number of cases and controls, incidence of the disease, confounders for which adjustment was made in multivariate analyses, type of effect estimate (odds ratio [OR], relative risk [RR], or incidence rate ratio [IRR]/hazard ratio [HR]), type of exposures examined and how they were measured (clinically or by laboratory analysis), fraction of the population with the exposure, and unadjusted and adjusted (at least for sexual behavior) effect estimate with 95% CI. Studies with incomplete data (e.g., no confidence intervals or measures of variation, or no description of the variables included in a multivariate analysis) were excluded from the analyses but included in the tables. When the effect estimate and a measure of variation could be calculated from the crude data in the report, these were included in the analysis. When several studies from the same population had been published, only the most recent were included in the meta-analysis. No quality indicator for the studies was derived to weight them.
Summary tables characterizing the studies and the results were constructed for the meta-analysis of studies investigating the effects of classic STDs on susceptibility to HIV. The tables were analyzed using fixed-model, inverse-variance meta-analysis. With this approach, the pooled-effect estimates are found by calculating a weighted average of the effects from each individual study, and the weight is the inverse of the variance (i.e., studies with narrow confidence intervals carry more weight). 17 This method is reliable when the studies are homogeneous (i.e., when they estimate the same effect that varies only with random error). This assumption can be tested by calculating the heterogeneity statistic, which is the weighted sum of the squared difference between the pooled estimate and the estimate from each individual study. This statistic follows a χ2 distribution. Publication bias was assessed by the Begg adjusted rank correlation test 18 and the Egger regression asymmetry test. 19 The statistical analyses were performed in STATA using the procedures meta and metabias. 20
Effects of STDs on Susceptibility to HIV
The literature search identified 30 longitudinal studies examining the effects of different STDs on susceptibility to HIV. Of these 30 studies, 17 also were covered in another recent review. 12 In terms of methods, 23 of the studies were prospective, either cohort 5,21–39 or nested case–control 16,40–43 studies, and 5 were retrospective, based on medical records. 44–48 One study had been published only as an abstract, 46 but more detailed information was given in a review, and the analyses were based on that. 12
Two studies were excluded from the analyses: one because of too few seroconversions, 37 the other because no proper effect estimates were reported. 48 The first study reported no association with any STDs, and the second found a higher HIV incidence among patients of STD clinics with GUDs than among those with NUDs. Three other studies, discussed later, were excluded from the initial analyses because there was no control group without an STD. 5,22,25
The data collected from all studies are summarized in Table 1. When several reports on the same study population were found (e.g., female sex workers in Kinshasa 23,40 and Mobasa 34,49 and military recruits in Chiang Mai 27,42), only the most recent information was used. Most of the studies examined the effects of several STDs at the same time. Of the 104 different effect estimates in the studies identified, 41 were crude estimates with no adjustments, and 63 were adjusted in multivariate models, at least for sexual behavior. When both unadjusted and adjusted effect estimates were reported, the adjusted estimate was used because it is important to control for confounding for sexual behavior. It is questionable then whether the unadjusted estimates also should be included from studies not reporting adjusted estimates for some or all of the STDs examined.
The problem of publication or reporting bias was relevant in this study. Some of the studies stated in the text that the effect of some STDs were insignificant without a report of the exact figures. Several of the studies excluded the STDs that did not give a significantly increased risk in univariate analyses from the final multivariate models, in which adjustment for sexual behavior was included. That is, they reported only unadjusted estimates for these STDs. The data were tested for publication bias by first including only the 63 adjusted estimates. The funnel plot depicted in Figure 2 indicates publication bias because it shows published effect estimates, with few small studies reporting relative risks below the pooled estimate (i.e., there is a tendency for publication only of studies showing a significant effect). This argument relies on the assumption that the identified studies have been drawn from a population of independent studies, and that studies with low power should be distributed evenly around the true effect. Both Begg’s test and Egger’s test for publication bias were highly significant (P < 0.01). Because the unadjusted estimates tended not to be significant, the problem of publication bias may have been overcome partly by inclusion also of studies reporting unadjusted estimates. When unadjusted effect estimates also were included, the tests suggested less bias, but still were significant (P = 0.03, Begg’s test;P < 0.01, Egger’s test). Both studies reporting adjusted and those reporting unadjusted estimates were included in the following analyses.
Meta-analyses for 11 different categories of STDs reported in the studies were conducted. The results are presented in Table 2 and Figure 3. The forest plots show the effect estimates from studies examining 10 different entities of STDs. The size of the boxes is inversely related to the size of the confidence intervals around the effect estimates, thereby reflecting the weight of the studies when the pooled estimates are calculated. The combined estimates are given for all the studies, and for the studies on only male-to-female or female-to-male transmission. As indicated in Table 2, for several of the STDs, the number of studies is small. Because of this and possible differences between study designs, calculated overall estimates are subject to uncertainty. Significant heterogeneity was found in the results for GUDs and NUDs, and also for candidiasis. Because the former two disease groups included several different STDs, and because the diagnoses were not laboratory confirmed, this heterogeneity is not surprising and suggests that the different STDs should not be grouped together. There possibly still is undetected heterogeneity in the studies of other individual STDs because the test has low power, and heterogeneity might be expected. This subject requires more extensive analyses.
Clinical GUDs together increase susceptibility to HIV, both in women and men, with the effect 1.6 times higher in men. Too few studies exist on any of the single GUDs (i.e., herpes, syphilis, and chancroid) for an examination of the difference between female and male susceptibility or a possible difference between the different GUDs. Although herpes and chancroid constitute most of the lesions in studies of clinical GUDs, few studies of these disorders exist, and almost all of these have used only serology for diagnosis (i.e., current lesions were not examined), except for one study of female chancroid in which culture was used.
Clinically diagnosed NUDs also have an effect in both women and men, with a relative difference between the genders similar to that found for GUDs. However, the effect is approximately 60% of that observed for GUDs. That GUDs have a stronger effect than NUDs also is shown by two studies in which the difference was examined directly. 5,22 When the results from these two studies were combined, the susceptibility to GUDs was shown to be significantly increased over susceptibility to NUDs (RR, 3.7; 95% CI, 1.6–8.7). Very few studies have investigated the effects that any of the specifically diagnosed NUDs have on male susceptibility.
Candidiasis and bacterial vaginosis both are a type of vaginitis and not traditionally considered STDs. 50 The presence of both depends on the microbial ecosystem of the vagina and may be linked to sexual activity. The Candida yeast and the bacteria prominent in bacterial vaginosis both may be transmitted from a male partner, but factors other than successful transmission are necessary, and these may also be sufficient for the establishment of vaginitis. Candidiasis seems to double the female susceptibility to HIV. The two studies on bacterial vaginosis indicate that clinical bacterial vaginosis increases susceptibility to HIV by a factor of 1.4 after adjustment. 33,35 A follow-up assessment of the latter study using the absence of lactobacilli or the presence of abnormal vaginal flora as indicators of bacterial vaginosis gave an effect estimate of 2. 34 This suggests that a vaginal milieu with a high pH and a lack of H2O2 production is favorable for HIV transmission. 51 Because bacterial vaginosis also is linked to an increased prevalence of other STDs, 52 it may be an important factor in HIV transmission.
Effects of STDs on HIV Infectiousness
Two longitudinal studies have investigated the effect of a concurrent STD on the HIV donor (i.e., HIV infectiousness) in discordant partnerships. In a study from Haiti, the unadjusted relative risk was 2.9 (95% CI, 1–9.1) when the donor had a GUD, 0.9 (95% CI, 0.1–6.9) when he or she had genital discharge, and 2.3 (95% CI, 1.1–4.6) when positive syphilis serology existed. 28 These effect estimates all were lower than those found in the HIV recipient with an STD: 6.8 (95% CI, 3–15.7), 2.6 (95% CI, 1.3–5), and 2.9 (95% CI, 1.4–6.2), respectively. These latter estimates possibly suggest a stronger effect of STDs on susceptibility than on infectiousness. The combined effect (i.e., STD in both donor and recipient) was 5.5 (95% CI, 0.8–36.6) for GUDs and 4.5 (95% CI, 1.3–15) for syphilis, indicating that the sum of increased susceptibility and infectiousness was less than additive.
A study from Rakai, Uganda, reported adjusted rate ratios of 1.6 (95% CI, 0.6–4.2) for GUD and 1.5 (95% CI, 0.7–3.3) for any STD in the HIV donor. 39 This was lower than the effect of GUD on susceptibility (RR, 3.1; 95% CI, 2–5) found in the parallel community cohort study, pointing also toward a lower effect on infectiousness than on susceptibility. Another report from the Rakai study found no effects of GUD, dysuria, syphilis, gonorrhea, chlamydia, trichomoniasis, or bacterial vaginosis, and a small but insignificant effect of genital discharge (RR, 1.9; 95% CI, 0.9–3.5). 53 This analysis was adjusted for blood viral load, and some STDs may increase both plasma viral load and genital shedding. Altogether, the data from these longitudinal studies are too limited for drawing firm conclusions.
Infectiousness has been studied indirectly by investigating the effect of STDs on biologic markers of increased transmission. The number of CD4 cells in endocervical specimens was significantly increased in patients with NUDs. 54 Chlamydia infections, gonorrhea, and bacterial vaginosis were associated with detection of cervical interleukin-10, which is an enhancer of macrophage HIV replication. 55 Both of these observations have been regarded as indicators of increased susceptibility, but they also could be markers of increased infectiousness. In women, HIV has been identified in GUDs, 56 whereas in men, it has been detected in chancroid ulcers 57 and herpes lesions. 58 In patients at STD clinics, HIV has been identified in herpes, syphilis, and chancroid ulcers. 59 In the latter study, detection of HIV was associated with a diagnosis of chancroid, a long-lasting ulcer and concurrent NUD. Detection of HIV in urethral specimens also is associated with urethritis and gonococcal infection. 60 Treatment of gonorrhea reduced the rate of HIV detection.
Several studies also have looked at the effect of STDs on seminal HIV. Two case reports showed that treatment of chlamydial urethritis 61 and gonococcal or nongonococcal urethritis 62 reduced seminal HIV. In a study comparing seminal HIV in patients with and those without urethritis, the viral load was approximately eight times higher in the men with urethritis. 63 It was higher in patients with gonorrhea than in those with nongonococcal urethritis, but both groups responded with a reduction after antibiotic treatment. Seminal viral load was reduced threefold after 2 weeks of treatment. On the basis of the reported linear correlation between blood and seminal viral load 64,65 and the 2.45 relative risk of seroconversion for each log increment in blood viral load, 53 the infectiousness caused by urethritis can be estimated as increased 1.5- to 2.2-fold, with the first estimate based on the effect from 2 weeks of treatment, and the second on the difference between the urethritis and control groups. This magnitude of effect accords with estimates from the discordant couple studies. Seminal viral load was increased in symptomatic but not in asymptomatic trichomonas urethritis. 63,66 Asymptomatic urethritis indicated by seminal leukocytosis was, however, associated with HIV seminal shedding. 67–69 Seminal viral load also was increased in patients with nongonococcal urethritis who had concomitant GUD. 70
Early studies found an association between cervical pus or inflammation and detection of cervical HIV. 71,72 A study of 223 pregnant women found a significant association between cervical HIV and cervical pus, but no association with gonorrhea or chlamydia. 73 An association between vaginal HIV and vaginal discharge also was observed, but no effect of trichomoniasis, candidiasis, or bacterial vaginosis. Another large study found that gonorrhea and cervical inflammation, but not chlamydia, were associated with endocervical HIV, whereas vaginal inflammation and candidiasis, but not trichomoniasis or bacterial vaginosis, were related to vaginal HIV shedding. 74 In a study of 604 women, detection of HIV in samples of cervicovaginal lavage was found to be associated with gonorrhea, chlamydia, and cervicovaginal ulcers, but not with trichomoniasis, syphilis, cervical pus, or vaginal discharge. 75 Although bacterial vaginosis was not related to increased HIV shedding in the aforementioned studies, a recently characterized factor that increases HIV replication has been found in cervicovaginal lavage samples. 76–78 This factor has been linked to bacterial vaginosis–associated, flora-like Mycoplasma hominis and Gardenerella vaginalis.79–81 Such a factor may be implicated in both increased infectiousness and susceptibility. On the basis of these qualitative and somewhat conflicting results, a quantitative assessment of the effects that STDs have on the infectiousness of women is not possible.
Effects of STDs on HIV Progression
The hypothesis that concurrent infections may alter the natural history of HIV is difficult to examine because a concurrent disease such as an opportunistic infection may be a marker and not a cause of progression. Considering that viral load is correlated with survival, the hypothesis that other infections alter progression to AIDS was supported by the findings that antigenic stimulation from tetanus toxoid, influenza, pneumococcal, and cholera vaccines transiently increased blood viral load. 82–86 The last of these vaccines was oral, suggesting that even a mucosal antigenic challenge (i.e., similar to some STDs) can induce viremia. Transient alterations in viral load also have been observed in patients with tuberculosis infection, bacterial pneumonia, and a variety of acute infections. 87–89
GUDs 90 have been associated with increased HIV viral load in cross-sectional studies, but that study design does not distinguish between possible effects of the STDs on HIV progression and effects of advanced disease on STD susceptibility and presentation. One longitudinal study found a transient increase in HIV viral load during active herpes simplex virus (HSV) infection. 91 Early studies suggested that coinfection with HIV and the other known pathogenic human retroviruses (i.e., human T-cell leukemia virus types 1 and 2) increased progression to AIDS. 92,93 However, human T-cell leukemia viruses do not increase HIV viral load, 94,95 and both a clinical study 96 and an animal model 97 showed no effect on survival.
In general, the evidence for an effect of transient increases in viremia on disease progression is lacking. A recent study found no effects of pertussis vaccination on disease progression in children infected with HIV. 98 Tuberculosis is the most studied intercurrent disease, and a review based on data from observational studies and randomized controlled trials concluded that there is no clear evidence for increased progression of HIV disease caused by intercurrent tuberculosis. 99 A meta-analysis has shown a significant beneficial effect of acyclovir treatment on HIV survival, suggesting that herpes simplex or varicella infections may increase disease progression. 100 However, firm evidence for an effect of intercurrent STDs on HIV progression is lacking, and an effect is not likely.
Effects of HIV on Susceptibility to STD
Relatively few epidemiologic studies have looked specifically at the effects of HIV on STD transmission. However, many of the cross-sectional case–control studies on the effects of STD on HIV transmission are equally relevant for examining the reciprocal relation because the odds ratios are the same if disease and exposure categories are interchanged. This was the main reason for not including cross-sectional studies in the meta-analysis of the effects that STDs have on susceptibility to HIV. The same arguments can be used for not relying on a cross-sectional design in studies of the opposite connection. Nevertheless, some cross-sectional studies are useful because the prevalence of STDs has been observed in different HIV disease stages based on clinical measures or CD4 counts.
The prevalence of chancroid ulcers in female sex workers in Kenya was greatest in women with generalized lymphadenopathy or AIDS who were HIV-positive. 101 A similar study of female sex workers from the Ivory Coast found a significant trend, with increased prevalences of GUDs, trichomoniasis, syphilis and genital warts, but not gonorrhea, in those who had HIV-positivity and decreasing CD4 counts. 102 Likewise, significant trends were found for bacterial vaginosis and abnormal vaginal discharge, but not for candidiasis and trichomoniasis, in women with HIV infection whose partners were male blood donors in Thailand who were HIV-positive. 103
These studies suggest a possible effect of HIV on susceptibility to several STDs, but a more rigorous analysis in longitudinal studies is required. Surprisingly few such studies have been conducted, although the incidence of STDs generally is higher than HIV incidence, which should make such studies easier than longitudinal studies of HIV transmission. Table 3 summarizes the five identified studies, 104–108 indicating that a significant effect of HIV is generally observed. The studies are too few for calculations of reliable estimates for the increased susceptibility. However, the two studies on herpes simplex infections together give a combined effect estimate of 4.4 (95 CI, 3.3–6).
Effects of HIV on STD Infectiousness
No epidemiologic studies investigating the effects of concurrent HIV and STD infection on the transmission of STDs were identified. Infectiousness evaluated by measuring the presence of bacteria or virus in genital samples has only been explored for herpes. Findings show that HSV-2 shedding can occur in both symptomatic and asymptomatic periods. 109 Such shedding was found to be four times more frequent in genital samples from women infected with HIV than in a control group. 110 Viral shedding was higher in women with lower CD4 counts. At delivery, HSV-2 shedding also is four times as common in pregnant women who are HIV-positive than in those who are HIV-negative. 111 A preliminary study on HSV-2 shedding among women from the Central African Republic found an increased prevalence and quantity of shedding in women who were HIV-positive, as compared with those who were HIV-negative. 112 Another recent study reported increased shedding in women with lowered CD4 counts. 113
In men who engage in sex with men, clinical anogenital HSV-2 shedding (i.e., shedding measured on a day of herpes symptoms) was as frequent in men who were HIV-negative as in men who were HIV-positive. However, subclinical shedding was six times as common in the men who were HIV-positive. 114 Altogether, it seems that concurrent HIV infection increases the asymptomatic shedding of HSV-2. How this will translate into an increased relative infectiousness of herpes is difficult to estimate.
Effects of HIV on STD Recovery and Recurrence
A relatively large literature exists concerning the effects of HIV infection on the natural history of classic STDs and their response to therapy. Most of this literature has limited value because they lack HIV-negative control groups, and few controlled studies were identified. Genital herpes recurrences are more common in men who are HIV-positive than in those who are HIV-negative (0.34 versus 0.23 recurrences per month or a ratio of 1.5). 114 Together with the increased asymptomatic shedding, this results in a markedly increased period of infectiousness. Prevalence of acyclovir-resistant HSV is higher in patients with HIV than in others. 115
A case–control study found that patients with HIV presented more often with secondary syphilis, and that chancres were more common. 116 This finding suggests a prolonged duration of infectiousness, but no prospective studies have confirmed this observation. There has been some debate on the effect of HIV on the serologic tests for syphilis. Nontreponemal tests such as the rapid plasma reagent test and the VDRL test seem to give more false positives in persons who are HIV-positive, whereas treponemal tests may serorevert in patients with HIV infection. 117–119 These altered laboratory tests probably have led to overestimation of treatment failures. A recent randomized controlled trial found few, and not a significantly different number of clinical treatment failures, in patients who were HIV-positive and those who were HIV-negative. 120
It has been reported that HIV infection causes decreased responsiveness to standard antibiotic treatment of chancroid. 121–124 More recent studies have found no effect of HIV serostatus on the number of treatment failures. 125–127 However, patients with HIV presented with ulcers of longer duration, and the time to healing was longer even after controlling for ulcer size. 125 One reason for the reported treatment failures may be increased prevalence of HSV-infected ulcers. 127
Studies examining the effect of HIV on the consequences of NUDs have focused on pelvic inflammatory disease. Although pelvic inflammatory disease seems to be more severe in women with HIV infection, treatment is successful. 128–135 When infected with chlamydia, patients who are HIV-positive have a higher risk for development of pelvic inflammatory disease than those who are HIV negative. 131 The bacterial etiologies of nonendometrial pelvic inflammatory disease (i.e., salpingitis) are similar between women with and those without HIV infection, 133 whereas HIV seems to increase the prevalence of endometritis with an atypical bacterial origin, possibly related to bacterial vaginosis. 134,135
Randomized controlled experimental studies have the strongest epidemiologic study design for establishing a causal relationship and for assessing intervention strategies. A causal effect of STDs on HIV transmission was found in a community intervention trial in Mwanza, Tanzania, 136 and this improved management of STDs was highly cost effective. 137 However, the next community randomized trial conducted in Rakai, Uganda, found no effect on HIV incidence. 138 The discrepancy between these two trials has been explained by differences in the stage of the HIV epidemic: greater effects of continuously improved treatment than of pulsed mass treatment, stronger impact of symptomatic than asymptomatic bacterial STDs, and differences in the prevalence of untreatable viral STDs. 139 A better and more quantitative understanding of the interactions between HIV infection and classic STDs is needed. Although the best indication of an STD effect on HIV transmission was provided by a randomized design, observational studies must be used to quantify the different interactions outlined in this systematic review.
Several problems are intrinsic to any systematic review and meta-analysis of observational studies in this field. In our review we identified three levels of bias. First, in each individual study, the fact that the route of transmission is common to the disease and the exposure of interest makes the quantification of association difficult. Sexual behavior is the common risk factor for contracting both HIV and STDs. Even after controlling for variables measuring number of partners, risk behavior, use of contraceptives, and the like, residual confounding probably will continue to exist because reliable and complete measures are lacking. Although the confidence intervals of our combined estimates are narrow, the underlying differences in study designs and adjustments to control for confounding and bias probably make them unreliable. They therefore represent a spurious precision. 13
Recently, systematic reviews on the impact of two other important potential confounders, female hormonal contraceptives and male circumcision, have been conducted. Two systematic reviews have examined the effects of hormonal contraceptives on HIV transmission. 140,141 Stephenson 140 concluded that the quality of the studies was too poor and the findings too inconsistent for a statistical meta-analysis, whereas Wang et al 141 performed a meta-analysis of more or less the same studies and found a significantly increased risk for HIV infection with use of oral contraceptives.
A meta-analysis of the studies examining the effect of circumcision on HIV transmission concluded, surprisingly, that men with circumcised penises had a small increased risk of HIV infection. 142 However, a reanalysis of the same set of studies with a proper statistical estimation of the overall effect reached the opposite conclusion. 143 Two other recent reviews concluded that there is substantial evidence for a protective effect of circumcision. 144,145 The latter review found evidence for a stronger protective effect of circumcision among high-risk men than among men in the general population. This raises the question whether the identified studies on the STD cofactor effect being conducted mostly in East Africa, and largely among high-risk groups are representative of the general population and other geographic regions. However, the studies, particular the prospective studies, must be conducted among groups in which the risk of infection is high enough to allow for robust estimation of the impact from risk factors.
These somewhat contradictory reviews also illustrate the problems of observational studies in this field. Examining the effects of STDs is still more complicated because STDs, in contrast to the two more constant factors, may be found in both the male and female partners, may be transmitted, and are dynamic entities in their own right. A particular problem is that the STD may be present for only a period of the time during which a person is at risk for HIV infection. This will tend to dilute the effect of the STD. 146 In addition, several of the studies relied on self-reported or serologically defined STDs, thereby suggesting that the STD may not have been present at the time of HIV transmission. Such a timing problem may lead further to underestimation of the effect.
Another problem is that both HIV and the STD may be transmitted at the same time. For example, in the study of Cameron et al, 5 the GUDs identified in male clients of female sex workers probably were caught from the women simultaneously with HIV, thereby suggesting that GUD increases infectiousness and not necessarily susceptibility. This would lead the study to conclude that the STD transmitted to the person at risk of HIV led to an increased susceptibility, whereas the effect may be explained solely by the impact of the STD on the infectiousness of the partner with HIV infection. This is an inherent problem in the reviewed studies.
In addition to the problem of confounding in the individual studies, the problem of misclassification also is evident. Several of the studies we identified relied on clinical diagnosis of the STD and therefore looked at the effects of heterogeneous entities such as GUDs and NUDs. Other studies used self-reported disease as a measurement of the exposure. Both approaches probably led to misclassification of the exposure. Hopefully, this misclassification was nondifferential, thereby leading generally to bias toward the null hypothesis of no association. Nevertheless, for robust studies, an etiologic laboratory diagnosis of the STD should be used. Too few studies have examined the individual STDs for firm conclusions to be drawn concerning the impact on any single STD.
The second level of bias involves reporting only adjusted effects from significant STDs in studies after controlling for confounders such as sexual behavior. Our review discovered that when a study identified an STD associated with a significantly increased risk of HIV infection, the study often reports only an adjusted effect estimate for this one STD and only unadjusted or no effect estimates for the other STDs examined. These unadjusted estimates probably are higher than they should be. As a result, the combined-effect estimates from meta-analyses would be overestimates.
Third, there is evidence of publication bias in our meta-analysis of studies concerning the effect of STDs on susceptibility to HIV. This likely bias is toward the reporting of significant results, which will cause the overall effect estimates to be overestimates. It probably is more difficult to identify potential unpublished work in observational epidemiology than in randomized trials, for which an effort is made to register all planned and ongoing trials. A recent assessment suggested that a publication bias existed in approximately half of a sample comprising systematic reviews of randomized controlled trials, but that it affected only the magnitude and not the quality of the conclusions (i.e., it had little bearing on the difference between an effect or no effect). 147
With all these biases, our estimates for the described interactions are uncertain. More information is needed on all the STDs and their interactions with HIV infection. It is possible on clinical grounds to group different STDs as ulcerative or nonulcerative, and as curable or incurable, although the observed heterogeneity suggests that the impact of different etiologic agents is not uniform. The limited empirical data reviewed are too restricted for detailed categories to be analyzed.
In a broad summary, assuming that confounding and publication biases were eliminated successfully, the studies indicated that GUDs increase male susceptibility approximately fourfold and female susceptibility about threefold, whereas NUDs increase male susceptibility threefold and female susceptibility twofold. There appears to be a twofold increase in the infectiousness of HIV from sources with an STD, although data are very sparse for women with an NUD. The combined effect of the HIV-susceptible and the infectious partner both having an STD, is greater than when one alone has the STD, but appears to be less than additive or multiplicative. This less than additive effect would be expected if the transmission probability became saturated when one of the partners was infected with the STD. Studies concerning the effects of HIV infection on STD transmission and duration were few. However, evidence was found for an impact of HIV on susceptibility to clinical GUDs and HSV, and on the shedding of HSV, although many more studies are required on this direction of the relation between HIV and STDs.
In general, the STDs seem to have a stronger effect on the susceptibility to HIV than on the infectiousness of HIV. This may suggest that interventions should be targeted to people not infected by HIV. However, the relative risks do not guide us to decide the importance of the increased risk on a population level. The population-attributable fraction or risk is a measure that takes the prevalence of the risk factor into account and describes the impact of the risk factor on the incidence of disease. Such a population level impact of the STDs on HIV incidence has been estimated. 16,39 Nonetheless, such a measure is problematic for infections. For instance, the effects on infectiousness may be more significant than the effects on susceptibility. Increased infectiousness will increase the transmission probability in all partnerships of the person infected with HIV and STD, whereas increased susceptibility will have an impact only in the partnerships where the person with STD infection engages in sex with a person who has HIV infection. The impact of STDs increasing HIV infectiousness may therefore be more important for the epidemiology of HIV even if it has a weaker effect, suggesting that treatment of STDs in patients who are HIV-positive should be targeted.
The baseline male-to-female HIV transmission probability is approximately twice the probability of transmission in the opposite direction. The data indicate that the susceptibility of men to HIV infection is more affected by the classic STDs than the susceptibility of women. Therefore, the HIV transmission probability from women to men is similar to the transmission probability from men to women in partnerships wherein the partner who is HIV negative also has an STD. This can have a major impact on the epidemiology of HIV because the likelihood of transmission from women to men may be a limiting factor. The gender differentials must, however, be confirmed by more studies in which an etiologic STD diagnosis is achieved because misclassification, especially that of NUDs in women, may lead to bias toward the null model, and because a gender differential in susceptibility is difficult to explain biologically.
Overall, the evidence points toward important positive bidirectional interactions between HIV infection and other STDs. The classic STDs and HIV infection may therefore reinforce the spread of each other and lead to a synergistic amplification, resulting in increased incidence and prevalence of both the STDs and the HIV infection. The impact of the epidemiologic interactions identified in our systematic review is, however, difficult to estimate in a static framework. Theoretical dynamic models of the combined epidemiology of STDs and HIV may aid such an understanding.
Despite our ability to estimate broadly the impact of STDs on susceptibility to HIV and infectiousness, and despite the large number of studies published, the major conclusion of this review must be that further studies with a more detailed focus are necessary before our understanding of this field is adequate. The meta-analysis on susceptibility to HIV has identified major problems related to reporting bias and problems in analysis of data. This should be kept in mind when new epidemiologic studies are designed and conducted.
First, the studies should allow for frequent STD sampling, thereby permitting its inclusion as a time-dependent covariate in survival analyses. Ideally, the number of sexual contacts with and without the STD should be taken into account when such studies are designed and analyzed because the transient nature of the STDs will dilute the relative risks found in epidemiologic studies, making the true increased risk per sexual contact much higher than the estimated risk. 146 Second, the data should be separated according to direction of heterosexual transmission because male susceptibility seems to increase more than female susceptibility. Third, sex behavior and other possible confounders must be measured in a rigorous way so adjustments can be made for them in the statistical analyses. Fourth, all STDs examined should be included in a multivariate analysis that allows control for confounders to reduce the problem of reporting bias. All this would suggest larger more detailed studies rather than a multiplicity of small often unreported studies. However, should the former prove to be prohibitively expensive, it may be worth establishing a database of studies examining risk factors similar to those internationally coordinated for studies estimating HIV prevalence.
Although our precise knowledge concerning the interaction of STDs and HIV is limited, we believe that STD control programs should be a part of all HIV prevention efforts. This conclusion stems in part from the promising results of the Mwanza project, 136 although they were not repeated in the Rakai trial, 138 and in part from the overall picture generated by this systematic review. Even if in the end it is found that STDs have only a limited impact on HIV transmission, we cannot afford to miss the potentially cost-effective chance of controlling HIV through their treatment. Additionally, STDs are important diseases, which by themselves cause major morbidity and reduced fertility, demanding control.
1. Piot P, Quinn TC, Taelman H, et al. Acquired immunodeficiency syndrome in a heterosexual population in Zaire. Lancet 1984; 2: 65–69.
2. Quinn TC, Glasser D, Cannon RO, et al. Human immunodeficiency virus infection among patients attending clinics for sexually transmitted diseases. N Engl J Med 1988; 318: 197–203.
3. Simonsen JN, Cameron DW, Gakinya MN, et al. Human immunodeficiency virus infection among men with sexually transmitted diseases: experience from a center in Africa. N Engl J Med 1988; 319: 274–278.
4. Greenblatt RM, Lukehart SA, Plummer FA, et al. Genital ulceration as a risk factor for human immunodeficiency virus infection. Aids 1988; 2: 47–50.
5. Cameron DW, Simonsen JN, D’Costa LJ, et al. Female-to-male transmission of human immunodeficiency virus type 1: risk factors for seroconversion in men. Lancet 1989; 2: 403–407.
6. Cameron DW, Padian NS. Sexual transmission of HIV and the epidemiology of other sexually transmitted diseases. Aids 1990; 4 (suppl 1): S99–S103.
7. Mertens TE, Hayes RJ, Smith PG. Epidemiological methods to study the interaction between HIV infection and other sexually transmitted diseases. Aids 1990; 4: 57–65.
8. Wasserheit JN. Epidemiological synergy: interrelationships between human immunodeficiency virus infection and other sexually transmitted diseases. Sex Transm Dis 1992; 19: 61–77.
9. Clottey C, Dallabetta G. Sexually transmitted diseases and human immunodeficiency virus: epidemiologic synergy? Infect Dis Clin North Am 1993; 7: 753–770.
10. Dickerson MC, Johnston J, Delea TE, White A, Andrews E. The causal role for genital ulcer disease as a risk factor for transmission of human immunodeficiency virus: an application of the Bradford Hill criteria. Sex Transm Dis 1996; 23: 429–440.
11. Dallabetta G, Diomi MC. Treating sexually transmitted diseases to control HIV transmission. Curr Opin Infect Dis 1997; 10: 22–25.
12. Fleming DT, Wasserheit JN. From epidemiological synergy to public health policy and practice: the contribution of other sexually transmitted diseases to sexual transmission of HIV infection. Sex Transm Infect 1999; 75: 3–17.
13. Egger M, Schneider M, Smith GD. Meta-analysis spurious precision? Meta-analysis of observational studies. BMJ 1998; 316: 140–144.
14. Thornton A, Lee P. Publication bias in meta-analysis: its causes and consequences. J Clin Epidemiol 2000; 53: 207–216.
15. Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis of Observational Studies in Epidemiology (MOOSE) group. JAMA 2000; 283: 2008–2012.
16. Orroth KK, Gavyole A, Todd J, et al. Syndromic treatment of sexually transmitted diseases reduces the proportion of incident HIV infections attributable to STD in rural Tanzania. AIDS 2000; 14: 1429–1437.
17. Deeks JJ, Altman DG, Bradburn MJ. Statistical methods for examining heterogeneity and combining results from several studies in meta-analysis. In: Egger M, Davey Smith GD, Altman DG, eds. Systematic Reviews in Health Care: Meta-analysis in Context. 2nd ed. London: BMJ Books, 2001.
18. Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics 1994; 50: 1088–1101.
19. Egger M, Smith GD, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997; 315: 629–634.
20. Sterne JAC, Bradburn MJ, Egger M. Meta-analysis in STATA. In: Egger M, Smith GD, Altman DG, eds. Systematic Reviews in Health Care: Meta-analysis in Context. 2nd ed. London: BMJ Books, 2001.
21. Plummer FA, Simonsen JN, Cameron DW, et al. Cofactors in male–female sexual transmission of human immunodeficiency virus type 1. J Infect Dis 1991; 163: 233–239.
22. Telzak EE, Chiasson MA, Bevier PJ, Stoneburner RL, Castro KG, Jaffe HW. HIV-1 seroconversion in patients with and without genital ulcer disease: a prospective study. Ann Intern Med 1993; 119: 1181–1186.
23. Laga M, Alary M, Nzila N, et al. Condom promotion, sexually transmitted diseases treatment, and declining incidence of HIV-1 infection in female Zairian sex workers. Lancet 1994; 344: 246–248.
24. 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.
25. Plourde PJ, Pepin J, Agoki E, et al. Human immunodeficiency virus type 1 seroconversion in women with genital ulcers. J Infect Dis 1994; 170: 313–317.
26. Mehendale SM, Rodrigues JJ, Brookmeyer RS, et al. Incidence and predictors of human immunodeficiency virus type 1 seroconversion in patients attending sexually transmitted disease clinics in India. J Infect Dis 1995; 172: 1486–1491.
27. Celentano DD, Nelson KE, Suprasert S, et al. Risk factors for HIV-1 seroconversion among young men in northern Thailand. JAMA 1996; 275: 122–127.
28. Deschamps MM, Pape JW, Hafner A, Johnson WD Jr. Heterosexual transmission of HIV in Haiti. Ann Intern Med 1996; 125: 324–230.
29. Mbizvo MT, Machekano R, McFarland W, et al. HIV seroincidence and correlates of seroconversion in a cohort of male factory workers in Harare, Zimbabwe. AIDS 1996; 10: 895–901.
30. Figueroa JP, Ward E, Morris J, et al. Incidence of HIV and HTLV-I infection among sexually transmitted disease clinic attenders in Jamaica. J Acquir Immun Defic Syndr Hum Retrovirol 1997; 15: 232–237.
31. Kapiga SH, Lyamuya EF, Lwihula GK, Hunter DJ. The incidence of HIV infection among women using family planning methods in Dar es Salaam, Tanzania. Aids 1998; 12: 75–84.
32. Nopkesorn T, Mock PA, Mastro TD, et al. HIV-1 subtype E incidence and sexually transmitted diseases in a cohort of military conscripts in northern Thailand. J Acquir Immun Defic Syndr Hum Retrovirol 1998; 18: 372–379.
33. Martin HL Jr, Nyange PM, Richardson BA, et al. Hormonal contraception, sexually transmitted diseases, and risk of heterosexual transmission of human immunodeficiency virus type 1. J Infect Dis 1998; 178: 1053–1059.
34. 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.
35. Taha TE, Hoover DR, Dallabetta GA, et al. Bacterial vaginosis and disturbances of vaginal flora: association with increased acquisition of HIV. AIDS 1998; 12: 1699–1706.
36. Kilmarx PH, Limpakarnjanarat K, Mastro TD, et al. HIV-1 seroconversion in a prospective study of female sex workers in northern Thailand: continued high incidence among brothel-based women. AIDS 1998; 12: 1889–1898.
37. Chirgwin KD, Feldman J, Dehovitz JA, Minkoff H, Landesman SH. Incidence and risk factors for heterosexually acquired HIV in an inner-city cohort of women: temporal association with pregnancy. J Acquir Immun Defic Syndr Hum Retrovirol 1999; 20: 295–299.
38. Rakwar J, Lavreys L, Thompson ML, et al. Cofactors for the acquisition of HIV-1 among heterosexual men: prospective cohort study of trucking company workers in Kenya. AIDS 1999; 13: 607–614.
39. Gray RH, Wawer MJ, Sewankambo NK, et al. Relative risks and population attributable fraction of incident HIV associated with symptoms of sexually transmitted diseases and treatable symptomatic sexually transmitted diseases in Rakai District, Uganda. Rakai Project Team. AIDS 1999; 13: 2113–2123.
40. Laga M, Manoka A, Kivuvu M, et al. Nonulcerative sexually transmitted diseases as risk factors for HIV-1 transmission in women: results from a cohort study [see comments]. AIDS 1993; 7: 95–102.
41. Weir SS, Feldblum PJ, Roddy RE, Zekeng L. Gonorrhea as a risk factor for HIV acquisition. AIDS 1994; 8: 1605–1608.
42. Nelson KE, Eiumtrakul S, Celentano D, et al. The association of herpes simplex virus type 2 (HSV-2), Haemophilus ducreyi,
and syphilis with HIV infection in young men in northern Thailand. J Acquir Immun Defic Syndr Hum Retrovirol 1997; 16: 293–300.
43. Bollinger RC, Brookmeyer RS, Mehendale SM, et al. Risk factors and clinical presentation of acute primary HIV infection in India. JAMA 1997; 278: 2085–2089.
44. Kassler WJ, Zenilman JM, Erickson B, Fox R, Peterman TA, Hook EW III. Seroconversion in patients attending sexually transmitted disease clinics. AIDS 1994; 8: 351–355.
45. Otten MW Jr, Zaidi AA, Peterman TA, Rolfs RT, Witte JJ. High rate of HIV seroconversion among patients attending urban sexually transmitted disease clinics. AIDS 1994; 8: 549–553.
46. Dominiguez K, Ellerbrock TV, Harrington PE, Bush T, Malecki J, Simonds RJ. Risk Factors for HIV Seroconversion Among Young Women in a Rural Community in the Southeastern United States. In: Program and abstracts of the XI International Conference on AIDS; July 7–12, 1996; Vancouver, British Columbia. Abstract Mo.C.222.
47. Weinstock H, Sweeney S, Satten GA, Gwinn M. HIV seroincidence and risk factors among patients repeatedly tested for HIV attending sexually transmitted disease clinics in the United States, 1991 to 1996. STD Clinic HIV Seroincidence Study Group. J Acquir Immun Defic Syndr Hum Retrovirol 1998; 19: 506–512.
48. Suligoi B, Giuliani M, Galai N, Balducci M. HIV incidence among repeat HIV testers with sexually transmitted diseases in Italy. STD Surveillance Working Group. AIDS 1999; 13: 845–850.
49. Martin JN, Ganem DE, Osmond DH, Page-Shafer KA, Macrae D, Kedes DH. Sexual transmission and the natural history of human herpesvirus 8 infection. N Engl J Med 1998; 338: 948–954.
50. Sobel JD. Vaginitis. N Engl J Med 1997; 337: 1896–1903.
51. Hillier SL. The vaginal microbial ecosystem and resistance to HIV. AIDS Res Hum Retroviruses 1998; 14 (suppl 1): S17–S21.
52. van De Wijgert JH, Mason PR, Gwanzura L, et al. Intravaginal practices, vaginal flora disturbances, and acquisition of sexually transmitted diseases in Zimbabwean Women. J Infect Dis 2000; 181: 587–594.
53. Quinn TC, Wawer MJ, Sewankambo N, et al. Viral load and heterosexual transmission of human immunodeficiency virus type 1. Rakai Project Study Group. N Engl J Med 2000; 342: 921–929.
54. Levine WC, Pope V, Bhoomkar A, et al. Increase in endocervical CD4 lymphocytes among women with nonulcerative sexually transmitted diseases. J Infect Dis 1998; 177: 167–174.
55. Cohen CR, Plummer FA, Mugo N, et al. Increased interleukin-10 in the endocervical secretions of women with nonulcerative sexually transmitted diseases: a mechanism for enhanced HIV-1 transmission? AIDS 1999; 13: 327–332.
56. Kreiss JK, Coombs R, Plummer F, et al. Isolation of human immunodeficiency virus from genital ulcers in Nairobi prostitutes. J Infect Dis 1989; 160: 380–384.
57. Plummer FA, Wainberg MA, Plourde P, et al. Detection of human immunodeficiency virus type 1 (HIV-1) in genital ulcer exudate of HIV-1-infected men by culture and gene amplification. J Infect Dis 1990; 161: 810–811.
58. Schacker T, Ryncarz AJ, Goddard J, Diem K, Shaughnessy M, Corey L. Frequent recovery of HIV-1 from genital herpes simplex virus lesions in HIV-1-infected men. JAMA 1998; 280: 61–66.
59. Gadkari DA, Quinn TC, Gangakhedkar RR, et al. HIV-1 DNA shedding in genital ulcers and its associated risk factors in Pune, India. J Acquir Immun Defic Syndr Hum Retrovirol 1998; 18: 277–281.
60. Moss GB, Overbaugh J, Welch M, et al. Human immunodeficiency virus DNA in urethral secretions in men: association with gonococcal urethritis and CD4 cell depletion. J Infect Dis 1995; 172: 1469–1474.
61. Eron JJ Jr, Gilliam B, Fiscus S, Dyer J, Cohen MS. HIV-1 shedding and chlamydial urethritis. JAMA 1996; 275: 36.
62. Atkins MC, Carlin EM, Emery VC, Griffiths PD, Boag F. Fluctuations of HIV load in semen of HIV-positive patients with newly acquired sexually transmitted diseases. BMJ 1996; 313: 341–342.
63. Cohen MS, Hoffman IF, Royce RA, et al. Reduction of concentration of HIV-1 in semen after treatment of urethritis: implications for prevention of sexual transmission of HIV-1. AIDSCAP Malawi Research Group. Lancet 1997; 349: 1868–1873.
64. Vernazza PL, Gilliam BL, Dyer J, et al. Quantification of HIV in semen: correlation with antiviral treatment and immune status. AIDS 1997; 11: 987–993.
65. Tachet A, Dulioust E, Salmon D, et al. Detection and quantification of HIV-1 in semen: identification of a subpopulation of men at high potential risk of viral sexual transmission. AIDS 1999; 13: 823–831.
66. Hobbs MM, Kazembe P, Reed AW, et al. Trichomonas vaginalis as a cause of urethritis in Malawian men. Sex Transm Dis 1999; 26: 381–387.
67. Anderson DJ, O’Brien TR, Politch JA, et al. Effects of disease stage and zidovudine therapy on the detection of human immunodeficiency virus type 1 in semen [see comments]. JAMA 1992; 267: 2769–2774.
68. Xu C, Politch JA, Tucker L, Mayer KH, Seage GR III, Anderson DJ. Factors associated with increased levels of human immunodeficiency virus type 1 DNA in semen. J Infect Dis 1997; 176: 941–947.
69. Winter AJ, Taylor S, Workman J, et al. Asymptomatic urethritis and detection of HIV-1 RNA in seminal plasma. Sex Transm Infect 1999; 75: 261–263.
70. Dyer JR, Gilliam BL, Eron JJ Jr, Cohen MS, Fiscus SA, Vernazza PL. Shedding of HIV-1 in semen during primary infection [letter]. AIDS 1997; 11: 543–545.
71. Clemetson DB, Moss GB, Willerford DM, et al. Detection of HIV DNA in cervical and vaginal secretions: prevalence and cor-relates among women in Nairobi, Kenya. JAMA 1993; 269: 2860–2864.
72. Kreiss J, Willerford DM, Hensel M, et al. Association between cervical inflammation and cervical shedding of human immunodeficiency virus DNA. J Infect Dis 1994; 170: 1597–1601.
73. John GC, Nduati RW, Mbori-Ngacha D, et al. Genital shedding of human immunodeficiency virus type 1 DNA during pregnancy: association with immunosuppression, abnormal cervical or vaginal discharge, and severe vitamin A deficiency. J Infect Dis 1997; 175: 57–62.
74. Mostad SB, Overbaugh J, DeVange DM, et al. Hormonal contraception, vitamin A deficiency, and other risk factors for shedding of HIV-1 infected cells from the cervix and vagina. Lancet 1997; 350: 922–927.
75. Ghys PD, Fransen K, Diallo MO, et al. The associations between cervicovaginal HIV shedding, sexually transmitted diseases, and immunosuppression in female sex workers in Abidjan, Cote d’Ivoire. AIDS 1997; 11: F85–F93.
76. Spear GT, al-Harthi L, Sha B, et al. A potent activator of HIV-1 replication is present in the genital tract of a subset of HIV-1-infected and uninfected women. AIDS 1997; 11: 1319–1326.
77. Al-Harthi L, Spear GT, Hashemi FB, Landay A, Sha BE, Roebuck KA. A human immunodeficiency virus (HIV)-inducing factor from the female genital tract activates HIV-1 gene expression through the KB enhancer. J Infect Dis 1998; 178: 1343–1351.
78. Hashemi FB, Spear GT, Madsen L, Mollenhauer J. Detection and molecular mass determination of an HIV replication-enhancing female genital tract factor using a blot bioassay [in process citation]. Biotechniques 2000; 28:478–480, 482, 484 passim.
79. Hashemi FB, Ghassemi M, Roebuck KA, Spear GT. Activation of human immunodeficiency virus type 1 expression by Gardnerella vaginalis.
J Infect Dis 1999; 179: 924–930.
80. Al-Harthi L, Roebuck KA, Olinger GG, et al. Bacterial vaginosis-associated microflora isolated from the female genital tract activates HIV-1 expression. J Acquir Immun Defic Syndr Hum Retrovirol 1999; 21: 194–202.
81. Olinger GG, Hashemi FB, Sha BE, Spear GT. Association of indicators of bacterial vaginosis with a female genital tract factor that induces expression of HIV-1. AIDS 1999; 13: 1905–1912.
82. Staprans SI, Hamilton BL, Follansbee SE, et al. Activation of virus replication after vaccination of HIV-1-infected individuals. J Exp Med 1995; 182: 1727–1737.
83. Ortigao-de-Sampaio MB, Shattock RJ, Hayes P, et al. Increase in plasma viral load after oral cholera immunization of HIV-infected subjects. AIDS 1998; 12: F145–F50.
84. Brichacek B, Swindells S, Janoff EN, Pirruccello S, Stevenson M. Increased plasma human immunodeficiency virus type 1 burden following antigenic challenge with pneumococcal vaccine. J Infect Dis 1996; 174: 1191–1199.
85. Stanley S, Ostrowski MA, Justement JS, et al. Effect of immunization with a common recall antigen on viral expression in patients infected with human immunodeficiency virus type 1 [see comments]. N Engl J Med 1996; 334: 1222–1230.
86. O’Brien WA, Grovit-Ferbas K, Namazi A, et al. Human immunodeficiency virus: type 1 replication can be increased in peripheral blood of seropositive patients after influenza vaccination. Blood 1995; 86: 1082–1089.
87. Goletti D, Weissman D, Jackson RW, et al. Effect of Mycobacterium tuberculosis
on HIV replication: role of immune activation. J Immunol 1996; 157: 1271–1278.
88. Bush CE, Donovan RM, Markowitz NP, Kvale P, Saravolatz LD. A study of HIV RNA viral load in AIDS patients with bacterial pneumonia. J Acquir Immun Defic Syndr Hum Retrovirol 1996; 13: 23–26.
89. Sulkowski MS, Chaisson RE, Karp CL, Moore RD, Margolick JB, Quinn TC. The effect of acute infectious illnesses on plasma human immunodeficiency virus (HIV) type 1 load and the expression of serologic markers of immune activation among HIV-infected adults. J Infect Dis 1998; 178: 1642–1648.
90. Dyer JR, Eron JJ, Hoffman IF, et al. Association of CD4 cell depletion and elevated blood and seminal plasma human immunodeficiency virus type 1 (HIV-1) RNA concentrations with genital ulcer disease in HIV-1-infected men in Malawi. J Infect Dis 1998; 177: 224–227.
91. Mole L, Ripich S, Margolis D, Holodniy M. The impact of active herpes simplex virus infection on human immunodeficiency virus load. J Infect Dis 1997; 176: 766–770.
92. Bartholomew C, Blattner W, Cleghorn F. Progression to AIDS in homosexual men coinfected with HIV and HTLV-I in Trinidad [letter]. Lancet 1987; 2: 1469.
93. Hattori T, Koito A, Takatsuki K, et al. Frequent infection with human T-cell lymphotropic virus type 1 in patients with AIDS but not in carriers of human immunodeficiency virus type 1. J Acquir Immun Defic Syndr 1989; 2: 272–276.
94. Ariyoshi K, Berry N, Wilkins A, et al. A community-based study of human immunodeficiency virus type 2 provirus load in rural village in West Africa. J Infect Dis 1996; 173: 245–248.
95. Harrison LH, Quinn TC, Schechter M. Human T-cell lymphotropic virus type 1 does not increase human immunodeficiency virus viral load in vivo. J Infect Dis 1997; 175: 438–440.
96. Hershow RC, Galai N, Fukuda K, et al. An international collaborative study of the effects of coinfection with human T-lymphotropic virus type 2 on human immunodeficiency virus type 1 disease progression in injection drug users. J Infect Dis 1996; 174: 309–317.
97. Fultz PN, McGinn T, Davis IC, Romano JW, Li Y. Coinfection of macaques with simian immunodeficiency virus and simian T cell leukemia virus type 1: effects on virus burdens and disease progression. J Infect Dis 1999; 179: 600–611.
98. Tovo PA, de Martino M, Gabiano C, Galli L. Pertussis immunization in HIV-1-infected infants: a model to assess the effects of repeated T-cell-dependent antigen administrations on HIV-1 progression. Italian Register for HIV Infection in Children. Vaccine 2000; 18: 1203–1209.
99. Del Amo J, Malin AS, Pozniak A, De Cock KM. Does tuberculosis accelerate the progression of HIV disease? Evidence from basic science and epidemiology. AIDS 1999; 13: 1151–1158.
100. Ioannidis JP, Collier AC, Cooper DA, et al. Clinical efficacy of high-dose acyclovir in patients with human immunodeficiency virus infection: a meta-analysis of randomized individual patient data. J Infect Dis 1998; 178: 349–359.
101. Cameron DW, Ngugi EN, Ronald AR, et al. Condom use prevents genital ulcers in women working as prostitutes: influence of human immunodeficiency virus infection. Sex Transm Dis 1991; 18: 188–191.
102. Ghys PD, Diallo MO, Ettiegne-Traore V, et al. Genital ulcers associated with human immunodeficiency virus–related immunosuppression in female sex workers in Abidjan, Ivory Coast [see comments]. J Infect Dis 1995; 172: 1371–1374.
103. Rugpao S, Nagachinta T, Wanapirak C, et al. Gynaecological conditions associated with HIV infection in women who are partners of HIV-positive Thai blood donors. Int J STD AIDS 1998; 9: 677–682.
104. Plummer FA, Simonsen JN, Chubb H, et al. Epidemiologic evidence for the development of serovar-specific immunity after gonococcal infection. J Clin Invest 1989; 83: 1472–1476.
105. Nasio JM, Nagelkerke NJ, Mwatha A, Moses S, Ndinya-Achola JO, Plummer FA. Genital ulcer disease among STD clinic attenders in Nairobi: association with HIV-1 and circumcision status. Int J STD AIDS 1996; 7: 410–414.
106. Kaul R, Kimani J, Nagelkerke NJ, et al. Risk factors for genital ulcerations in Kenyan sex workers: the role of human immunodeficiency virus type 1 infection. Sex Transm Dis 1997; 24: 387–392.
107. Kamali A, Nunn AJ, Mulder DW, Van Dyck E, Dobbins JG, Whitworth JA. Seroprevalence and incidence of genital ulcer infections in a rural Ugandan population. Sex Transm Infect 1999; 75: 98–102.
108. McFarland W, Gwanzura L, Bassett MT, et al. Prevalence and incidence of herpes simplex virus type 2 infection among male Zimbabwean factory workers. J Infect Dis 1999; 180: 1459–1465.
109. Wald A, Zeh J, Selke S, et al. Reactivation of genital herpes simplex virus type 2 infection in asymptomatic seropositive persons. N Engl J Med 2000; 342: 844–850.
110. Augenbraun M, Feldman J, Chirgwin K, et al. Increased genital shedding of herpes simplex virus type 2 in HIV-seropositive women. Ann Intern Med 1995; 123: 845–847.
111. Hitti J, Watts DH, Burchett SK, et al. Herpes simplex virus seropositivity and reactivation at delivery among pregnant women infected with human immunodeficiency virus-1. Am J Obstet Gynecol 1997; 177: 450–454.
112. Mbopi Keou FX, Gresenguet G, Mayaud P, et al. Genital herpes simplex virus type 2 shedding is increased in HIV-infected women in Africa. Aids 1999; 13: 536–537.
113. Mostad SB, Kreiss JK, Ryncarz AJ, et al. Cervical shedding of herpes simplex virus in human immunodeficiency virus–infected women: effects of hormonal contraception, pregnancy, and vitamin A deficiency. J Infect Dis 2000; 181: 58–63.
114. Schacker T, Zeh J, Hu HL, Hill E, Corey L. Frequency of symptomatic and asymptomatic herpes simplex virus type 2 reactivations among human immunodeficiency virus–infected men. J Infect Dis 1998; 178: 1616–1622.
115. Severson JL, Tyring SK. Relation between herpes simplex viruses and human immunodeficiency virus infections. Arch Dermatol 1999; 135: 1393–1397.
116. Hutchinson CM, Hook EW III, Shepherd M, Verley J, Rompalo AM. Altered clinical presentation of early syphilis in patients with human immunodeficiency virus infection. Ann Intern Med 1994; 121: 94–100.
117. Telzak EE, Greenberg MS, Harrison J, Stoneburner RL, Schultz S. Syphilis treatment response in HIV-infected individuals. Aids 1991; 5: 591–595.
118. Janier M, Chastang C, Spindler E, et al. A prospective study of the influence of HIV status on the seroreversion of serological tests for syphilis. Dermatology 1999; 198: 362–369.
119. Joyanes P, Borobio MV, Arquez JM, Perea EJ. The association of false-positive rapid plasma reagin results and HIV infection. Sex Transm Dis 1998; 25: 569–571.
120. Rolfs RT, Joesoef MR, Hendershot EF, et al. A randomized trial of enhanced therapy for early syphilis in patients with and without human immunodeficiency virus infection. The Syphilis and HIV Study Group. N Engl J Med 1997; 337: 307–314.
121. MacDonald KS, Cameron DW, D’Costa L, Ndinya-Achola JO, Plummer FA, Ronald AR. Evaluation of fleroxacin (RO 23–6240) as single-oral-dose therapy of culture-proven chancroid in Nairobi, Kenya. Antimicrob Agents Chemother 1989; 33: 612–614.
122. Tyndall M, Malisa M, Plummer FA, Ombetti J, Ndinya-Achola JO, Ronald AR. Ceftriaxone no longer predictably cures chancroid in Kenya. J Infect Dis 1993; 167: 469–471.
123. Tyndall MW, Agoki E, Plummer FA, Malisa W, Ndinya-Achola JO, Ronald AR. Single-dose azithromycin for the treatment of chancroid: a randomized comparison with erythromycin. Sex Transm Dis 1994; 21: 231–234.
124. Behets FM, Liomba G, Lule G, et al. Sexually transmitted diseases and human immunodeficiency virus control in Malawi: a field study of genital ulcer disease. J Infect Dis 1995; 171: 451–455.
125. King R, Choudhri SH, Nasio J, et al. Clinical and in situ cellular responses to Haemophilus ducreyi
in the presence or absence of HIV infection. Int J STD AIDS 1998; 9: 531–536.
126. Bogaerts J, Kestens L, van Dyck E, Tello WM, Akingeneye J, Mukantabana V. Genital ulcers in a primary health clinic in Rwanda: impact of HIV infection on diagnosis and ulcer healing (1986–1992). Int J STD AIDS 1998; 9: 706–710.
127. Malonza IM, Tyndall MW, Ndinya-Achola JO, et al. A randomized, double-blind, placebo-controlled trial of single-dose ciprofloxacin versus erythromycin for the treatment of chancroid in Nairobi, Kenya. J Infect Dis 1999; 180: 1886–1893.
128. Hoegsberg B, Abulafia O, Sedlis A, et al. Sexually transmitted diseases and human immunodeficiency virus infection among women with pelvic inflammatory disease. Am J Obstet Gynecol 1990; 163: 1135–1139.
129. Korn AP, Landers DV, Green JR, Sweet RL. Pelvic inflammatory disease in human immunodeficiency virus–infected women. Obstet Gynecol 1993; 82: 765–768.
130. Kamenga MC, De Cock KM, St. Louis ME, et al. The impact of human immunodeficiency virus infection on pelvic inflammatory disease: a case–control study in Abidjan, Ivory Coast. Am J Obstet Gynecol 1995; 172: 919–925.
131. Kimani J, Maclean IW, Bwayo JJ, et al. Risk factors for Chlamydia trachomatis
pelvic inflammatory disease among sex workers in Nairobi, Kenya. J Infect Dis 1996; 173: 1437–1444.
132. Barbosa C, Macasaet M, Brockmann S, Sierra MF, Xia Z, Duerr A. Pelvic inflammatory disease and human immunodeficiency virus infection. Obstet Gynecol 1997; 89: 65–70.
133. Cohen CR, Sinei S, Reilly M, et al. Effect of human immunodeficiency virus type 1 infection upon acute salpingitis: a laparoscopic study. J Infect Dis 1998; 178: 1352–1358.
134. Bukusi EA, Cohen CR, Stevens CE, et al. Effects of human immunodeficiency virus 1 infection on microbial origins of pelvic inflammatory disease and on efficacy of ambulatory oral therapy. Am J Obstet Gynecol 1999; 181: 1374–1381.
135. Irwin KL, Moorman AC, O’Sullivan MJ, et al. Influence of human immunodeficiency virus infection on pelvic inflammatory disease. Obstet Gynecol 2000; 95: 525–534.
136. Grosskurth H, Mosha F, Todd J, et al. Impact of improved treatment of sexually transmitted diseases on HIV infection in rural Tanzania: randomised controlled trial. Lancet 1995; 346: 530–536.
137. Gilson L, Mkanje R, Grosskurth H, et al. Cost effectiveness of improved treatment services for sexually transmitted diseases in preventing HIV-1 infection in Mwanza Region, Tanzania [see comments]. Lancet 1997; 350: 1805–1809.
138. Wawer MJ, Sewankambo NK, Serwadda D, et al. Control of sexually transmitted diseases for AIDS prevention in Uganda: a randomised community trial. Rakai Project Study Group. Lancet 1999; 353: 525–535.
139. Grosskurth H, Gray R, Hayes R, Mabey D, Wawer M. Control of sexually transmitted diseases for HIV-1 prevention: understanding the implications of the Mwanza and Rakai trials. Lancet 2000; 355: 1981–1987.
140. Stephenson JM. Systematic review of hormonal contraception and risk of HIV transmission: when to resist meta-analysis. AIDS 1998; 12: 545–553.
141. Wang CC, Kreiss JK, Reilly M. Risk of HIV infection in oral contraceptive pill users: a meta-analysis. J Acquir Immun Defic Syndr Hum Retrovirol 1999; 21: 51–58.
142. Van Howe RS. Circumcision and HIV infection: review of the literature and meta-analysis. Int J STD AIDS 1999; 10: 8–16.
143. O’Farrell N, Egger M. Circumcision in men and the prevention of HIV infection: a “meta-analysis” revisited. Int J STD AIDS 2000; 11: 137–142.
144. Moses S, Bailey RC, Ronald AR. Male circumcision: assessment of health benefits and risks. Sex Transm Infect 1998; 74: 368–373.
145. Weiss HA, Quigley MA, Hayes RJ. Male circumcision and risk of HIV infection in sub-Saharan Africa: a systematic review and meta-analysis. AIDS 2000; 14: 2361–2370.
146. Hayes RJ, Schulz KF, Plummer FA. The cofactor effect of genital ulcers on the per-exposure risk of HIV transmission in sub-Saharan Africa. J Trop Med Hyg 1995; 98: 1–8.
147. Sutton AJ, Duval SJ, Tweedie RL, Abrams KR, Jones DR. Empirical assessment of effect of publication bias on meta-analyses. BMJ 2000; 320: 1574–1577.