Sample EPID168 commentary (Fall 1997)
Gonorrhea and other STD as risk factors for the transmission of HIV in the United States
Since early in the HIV epidemic, the role of other sexually transmitted diseases (STD) as risk factors for HIV transmission has been a question with important implications for the control of the spread of HIV infection. If other STD are important cofactors in HIV transmission, then control of these diseases, particularly in areas with high prevalence of both STD and HIV, would be a simple and relatively inexpensive public health measure to slow the HIV epidemic. This paper examines research on this potential association, with particular focus on studies conducted in the United States.
STD could increase susceptibility to HIV infection in several ways. First, the ulcerations associated with Herpes Simplex Virus (HSV), syphilis, and chancroid compromise the skin's physical barrier to infection and provide an infectious agent easier access to the bloodstream. Second, STD infections could have an immunosuppresive effect and make someone more susceptible to HIV infection. Third, STD infection causes a recruitment of target cells to HIV, lymphocytes and macrophages, into the genital mucosa making them accessible to HIV infection.1 These infected cells can then cross back into the bloodstream and disseminate the infection.
Other STD may also facilitate HIV transmission by increasing the infectivity of persons co-infected with HIV by 1) recruiting HIV infected lymphocytes or macrophages to the genital region or 2) increasing the levels of HIV virions present in genital secretions.2 This increase in viral burden could be due to increased HIV replication brought on by immune activation in response to infection with other STD. Increased viral load may also occur by direct interaction between HIV and another viral STD. Studies in vitro3 and in vivo4 of HIV-1 infection have shown that coinfection with herpes viruses can lead to increased HIV replication via a synergistic activation of the viral promoter regions in the two viruses.
Much of the early research examining the role of STD on HIV transmission was done in central Africa on commercial sex workers and among people attending STD clinics. Overall, these studies provide strong evidence that HIV transmission is enhanced by ulcerative STDs5 and by non-ulcerative STDs such as gonorrhea and chlamydia6 and trichomoniasis.5,6 While such studies are important from a global perspective, they have limited relevance in the United States since the African HIV epidemic is spread largely through heterosexual contact whereas the US epidemic is found predominantly in homosexual males and injection drug users. Also, the distribution of non-HIV STDs differs substantially between Africa and North America.
Studies in the US have been conducted primarily in homosexual men and in people attending STD clinics. Despite the differences in the study populations and epidemiologic methods used, the evidence that STD enhance HIV transmission is fairly consistent. A case-control study of 179 HIV infected patients and 367 age-, race- and gender-matched HIV seronegative controls conducted in STD clinics in Baltimore found that history of HSV-2 infection by antibody testing was significantly associated with HIV infection among heterosexual men and women.7 A nested case-control study of 62 incident HIV infections and 61 seronegative controls in a cohort of homosexual men in San Francisco showed that prior infection with HSV-2 was a risk factor for HIV infection.8 One of the few prospective studies conducted in the US was a cohort of 1669 initially HIV seronegative New York City STD clinic patients who were followed for a minimum of 3 months. A multiple logistic regression model controlling for known HIV risk factors showed that in the heterosexual male group who completed the study (N=758), chancroid (OR = 3.3) and acquisition of STD during follow up (OR = 3.2) were significantly associated with HIV seroconversion.9
While the above studies have shown an association between STD and transmission of HIV in the United States, there are weaknesses in these studies. A major methodologic concern for all of the studies is that because HIV is spread primarily by sexual contact, it shares many risk factors associated with other STD, and this makes it difficult to control for the confounding with these other risk factors. It is possible that the association between STD and HIV transmission may due to exposure to sexual partners coinfected with both and have little to do with the other STD infections. A second weakness common to these studies is that the sexual behavior data were self-reported and are thus subject to recall bias and frankness of response. Finally, unprotected sexual contact with HIV infected partners was not obtained. This is the "true" risk factor for HIV infection, and it is probably impossible to obtain an accurate assessment of this given the long latency between HIV infection and symptoms. As a result, all of these studies assume that number of partners and sexual practices are unbiased surrogate markers for contact with HIV infected partners, and while this assumption is reasonable, some misclassification of this exposure is likely in particular because risk reduction through condom use was not examined by any of these studies.
In addition to these common weaknesses, the studies cited have specific problems which must be considered. The case-control study by Hook showed an association between "prior" HSV infection and HIV infection,7 but it is impossible to tell whether the HSV infection preceded the HIV infection or was concurrent with it, and this lack of temporal sequence is the most important drawback to the case-control design. In addition, in a multiple regression model used in this study, injection drug use was a much stronger risk factor (OR = 6.0, p < 0.001) than was HSV seropositivity (OR = 1.9, p = 0.024). This result is expected given the relatively high prevalence of injection drug use and suggests that HSV is not the most important HIV risk factor in this population.
The prospective study by Telzak had a very short follow up time, with a stated minimum of 3 months and no data on the overall follow up of the cohort.9 As a consequence, the number of incident HIV infections was small (10 of 344 in the group with ulcerative STD and 4 of 414 in the comparison group), and the relative rates of infection are subject to instability due to small numbers.
The nested case-control study by Holmberg8 is the best of the ones reviewed here. Their matching of controls to cases was done by year of serum sample collection which is in essence a time density sampling paradigm. Thus, the study should not be subject to secular trends in infection rates, sexual behaviors, HIV and STD incidence or treatments, and the behavioral data should not be subject to recall bias. Also, by using sera collected prior to and after HIV infection, they were able to show that HSV infection preceded HIV infection. Unfortunately, the number of participants in this study was also small and subject to low power and stability of estimates.
Conclusions and recommendations
This limited survey of the interaction between STD and HIV has found that STDs are risk factors for HIV transmission. The evidence is strong and consistent for genital ulcer diseases and is generally convincing for non-ulcerative STDs. Research on this topic has focused on HSV, gonorrhea, syphilis, and chancroid. Conspicuously understudied are infectious agents such as human papilloma virus, Candida, trichomonas, and non-specific genital inflammations such as bacterial vaginosis. These are common STDs and some effort should be made in examining any role they may have as risk factors for HIV transmission.
Clearly, efforts to control STD other then HIV would be helpful in slowing the HIV epidemic and efforts to control these infections should be implemented. However, it is also important to remember that STD infections are responsible for significant morbidity and mortality on their own and should be targeted for control on this basis also, and not just in the context of their influence on HIV infection.