Since the beginning of the HIV epidemic, HIV seroincidence has remained higher among men who have sex with men (MSM) than in any other risk group in the United States.1,2 Despite decreases in HIV seroincidence accompanying substantial declines in risk behavior in the 1980s, recent data suggest increases in HIV risk factors3-5 and HIV seroincidence6,7 in the United States, possibly fueled by beliefs in the efficacy of antiretroviral therapy in reducing HIV transmission and disease progression.6,8-10
Most studies evaluating risk factors for prevalent11-15 or incident16-19 HIV infection were conducted during the first 10 years of the epidemic when risk behavior patterns appeared to be quite different from those reported in later years. Limited data exist on risk factors for incident HIV infection in more recent years.20,21 One study of MSM enrolled in 3 US cities into a Centers for Disease Control and Prevention (CDC)-sponsored HIV vaccine preparedness cohort in the early 1990s found that age, unprotected receptive anal sex (VRA), injection drug use, urethritis, and condom failure were independent risk factors for infection.20 The current study extends these findings by analyzing data from a later substantially larger HIV vaccine preparedness cohort that enrolled MSM from 6 US cities at the time more potent antiretroviral therapy became available.
The HIV Network for Prevention Trials (HIVNET) conducted the Vaccine Preparedness Study (VPS), a 9-city prospective cohort study of high-risk MSM, women at heterosexual risk, and injection drug users from April 1995 through May 1997.22 This analysis is limited to the 3257 MSM enrolled in 6 cities: Boston, Chicago, Denver, New York, San Francisco, and Seattle. Participants were seen every 6 months over an 18-month period for HIV antibody pretest counseling and testing and behavioral interviews; HIV posttest counseling occurred at separate visits 2 weeks later. The common protocol was approved by all local institutional review boards, and all participants gave written informed consent before participation in the study.
Participants were eligible to participate if they were negative for HIV antibodies on enrollment and reported having anal sex (protected or unprotected) with a man in the 12 months before enrollment. Men were recruited from prior cohort studies, sexually transmitted disease clinics, bars and dance clubs, advertising, street outreach, and referral from other participants. Behavioral assessments were conducted by trained interviewers using a standardized instrument and included questions about demographics, self-perceived risk of acquiring HIV infection, number of penetrative sex partners (anal, vaginal, or oral), sexual behaviors stratified by partner HIV serostatus, drug and alcohol use, recent clinical symptoms, self-reported sexually transmitted infections and circumcision status, and attitudes toward participation in future prevention trials; copies of the questionnaire are available on request. All questionnaires assessed behaviors that occurred in the prior 6-month period.
Multivariable “pooled” logistic regression models were used to identify the independent predictors of HIV seroconversion. In this model, a dichotomous seroconversion outcome is defined for each subject in each 6-month interval in which he remains at risk. The pooled outcomes are analyzed using a logistic regression model that includes a separate intercept for each interval. Like the relative hazards estimated using Cox models, adjusted odds ratios (AORs) estimated using the pooled logistic regression model are interpretable as the relative odds of seroconversion in the interval between the previous and current visits, given that this event has not previously occurred. The final multipredictor model using time-dependent covariates was found using backward deletion of predictor variables significant at P < 0.2 in single-predictor models. This selection procedure was robust and verified in alternative models, including models in which numbers of contacts were taken into account. Population attributable risk (PAR) was estimated using the fitted risks from the pooled logistic regression models.23 PAR estimates for individual predictors based on a multiplicative model may sum to more than 100% and thus are best interpreted as estimating the relative importance of various risk factors. Analyses were carried out using S-plus, version 6.1 (Mathsoft, Seattle, WA).
The demographics of the enrolled cohort are shown in Table 1. Among the 3257 MSM enrolled in the HIVNET VPS, 72 became infected with HIV during 8848 visits throughout the study. The annualized HIV seroincidence was 1.55 per 100 person-years (95% confidence interval [CI]: 1.23-1.95), with no statistically significant decline during the 18 months of follow-up (seroincidence at 6-month visit = 1.66, 95% CI: 1.13-2.43; seroincidence at 12-month visit = 1.81, 95% CI: 1.24-2.64; and seroincidence at 18-month visit = 1.15, 95% CI: 0.73-1.83). Of 3257 men, 393 (12%) failed to complete the study.
On univariate analysis, a number of demographic and risk factors were significantly associated with HIV seroconversion (Table 2). Younger men, less educated men, and those lacking private health insurance were more likely to become infected with HIV. Men who were uncircumcised, those with a larger number of sexual partners, those engaging in any of several types of unprotected sex with HIV-positive or unknown serostatus partners, those reporting protected receptive anal sex (PRA), and those engaging in group sex or public sex were also at greater risk than those not reporting these risks on univariate analysis. Men with a self-reported history of urethral or rectal gonorrhea or Chlamydia were also at increased risk. Finally, several drugs were also associated with HIV seroconversion, including nitrite inhalants (“poppers”), amphetamines, cocaine, hallucinogens, and injection drugs.
On multivariate analysis (Table 3), after controlling for specific sexual behaviors, the only demographic variable that remained significantly associated with HIV seroconversion was lack of private health insurance. The association of younger age with seroconversion approached but did not achieve statistical significance; the association was weaker when age was stratified more finely and analyzed as either ordinal or nominal. Uncircumcised men were approximately twice as likely as circumcised men to seroconvert. A number of sexual risk variables were independently associated with an approximate doubling or tripling of risk of HIV seroconversion, including URA with an HIV-positive or unknown serostatus partner, PRA with an HIV-positive partner, and unprotected receptive oral sex with ejaculation (UOE) with an HIV-positive partner. The risk of seroconversion increased by 14% with each additional HIV-negative sex partner in the prior 6 months. Nitrite inhalant use was independently associated with more than a doubling of risk.
We performed supplemental analyses to explore possible causal pathways by which some of these variables may affect the risk of HIV seroconversion. There was no significant interaction between circumcision and self-reported sexually transmitted diseases (AOR for interaction = 2.2, 95% CI: 0.6-8.9) or unprotected insertive anal sex (UIA; AOR for interaction = 0.8, 95% CI: 0.3-2.4) or interaction with other demographic data (eg, city of recruitment, race/ethnicity, age) or substance use (eg, poppers). Similarly, there was no significant interaction between nitrite inhalant use and specific sexual practices, sexually transmitted diseases, or the total number of sex partners (data not shown).
Because the risk associated with either UOE or PRA was previously thought to be relatively low, we further describe the other risk behaviors reported by seroconverters engaging in either of these activities. Of 7 seroconverters reporting UOE with an HIV-positive partner, 6 reported other unprotected sexual contacts that have been estimated to carry a higher per-contact risk of HIV acquisition.24 Of these 6 individuals, 3 reported URA with an HIV-positive partner, 2 others reported UIA with an HIV-positive partner, and 1 reported URA only with HIV-negative partners but reported 15 HIV-negative partners. The remaining participant reported only PRA with an HIV-positive partner.
Of the 16 seroconverters reporting PRA with an HIV-positive partner, only 3 reported no unprotected risk behaviors. Of the remaining 13 individuals, 9 reported URA with an HIV-positive partner, 2 reported URA with an unknown serostatus partner, 2 reported UIA with an HIV-positive partner, 1 reported UIA only with HIV-negative partners but reported 10 HIV-negative partners, and 1 reported only UOE with an HIV-positive partner.
The PAR associated with having at least 1 of the independent risk factors for seroconversion was 82.3% (see Table 3). Having a larger number of reportedly HIV-negative sex partners and nitrite inhalant use carried the greatest PAR, with each potentially accounting for more than one quarter of new infections. URA with an HIV-positive or unknown serostatus partner carried a PAR of 12% and 15%, respectively, with a combined PAR of 25%. PRA with an HIV-positive partner and lack of circumcision were each associated with a PAR of approximately 10%. UOE with an HIV-positive partner carried a PAR of only 7%.
The risk factors appearing to contribute to the greatest number of infections in this cohort included a larger number of HIV-negative sex partners, inhaled nitrite use, and younger age. Although many counseling messages aimed at HIV-negative persons focus on unsafe sexual activity with HIV-positive sex partners or those of unknown HIV serostatus, our study suggested that more than one quarter of new infections arose from men having HIV-“negative” partners. This probably reflects error in presumed partner serostatus, particularly among men with multiple partners, as well as residual confounding by factors such as partner infectiousness and level of sexual trauma. There is an increasing literature on the importance of understanding patterns of sexual mixing on HIV spread,25,26 because the rate of sexual partner change is an important variable in the equation of epidemic spread of infectious diseases.27 Recent data suggest that as many as 25% of the HIV-infected population in the United States28 and up to 77% of young HIV-infected MSM29 are unaware of their HIV serostatus and that 42% of MSM failed to disclose HIV serostatus with at least 1 partner, particularly in nonexclusive partnerships.30 Recent licensure of rapid HIV tests may provide opportunities to increase awareness and disclosure of current HIV serostatus in high-risk populations. Prevention messages should emphasize the importance of reducing the number of sexual partners, even those presumed to be HIV-negative.
Nitrite inhalers have long been reported as associated with prevalent HIV infection13,17 and high-risk sexual practices31-34 and were suspected of being the causal agent of AIDS in the early years of the epidemic. Although the independent contribution of nitrite inhalants to incident infection has been inconsistent across studies, our study and others16,35 found a significant association of nitrite inhalants with prospectively identified HIV infection. Recently, nitrite inhalants have also been found to be independently associated with Kaposi sarcoma-associated herpesvirus (KSHV) infection acquisition.36 The pathway by which nitrite inhalants lead to HIV or KSHV infection are not clear. Nitrites inhalants cause peripheral vasodilatation and are believed to decrease anal sphincter tone, potentially leading to more traumatic sexual intercourse or more direct exposure to blood cells.37 Because inhaled nitrites are generally used expressly to increase sexual pleasure, their use may be associated with other unmeasured confounders, including increased partner infectiousness or level of sexual trauma. There are also limited animal38 and human data39 suggesting that nitrite inhalants may cause transient immunosuppression or alter cytokine profiles, which could enhance transmission HIV or KSHV transmission across mucosal barriers. The high prevalence of nitrite inhalant use among highrisk MSM, recently reported by 37% of MSM enrolled in a large multisite behavioral intervention34 and in 16% to 21% of a multisite household-based sample of urban MSM,40 reinforces the need to develop prevention strategies to decrease nitrite inhalant use or alter sexual practices when these substances are used. The latter alternative may be particularly challenging, because nitrite inhalants are used expressly to enhance sexual pleasure. Prevention strategies may need to be adapted to include other recreational drugs as patterns of use change over time.
Recent cross-sectional studies suggest high levels of risk practices and projected HIV seroincidence among men less than 25 years of age7; a prospective seroincidence study conducted in the early 1990s found young age to be an independent predictor of HIV seroconversion.20 Possible mechanisms for increased acquisition among younger men include higher rates of condom failure,41 more frequent misclassification of partner serostatus, and unmeasured variables associated with sexual risk such as trauma during initiation of sexual practices or exposure to recently infected and thus highly infectious cases. In our study, risk estimates were similar in the age groups <25 years old and 26 to 35 years old. Although age ≤35 years did not reach formal statistical significance, its estimated PAR was nominally largest of all the risk factors included in the multipredictor model. Therefore, prevention efforts should be developed that focus particularly on young men, recognizing that this increased risk appears to extend to men in their mid-30s. Our study failed to find an independent association of race/ethnicity with HIV seroconversion as has been reported in recent studies of African-American and Latino MSM,5,42-45 although our power to detect such differences may have been limited by the small numbers of African-American and Latino men enrolled in the study.
Although 2 meta-analyses of the role of circumcision in sub-Saharan Africa have demonstrated a reduced risk of HIV infection in circumcised heterosexual men,46,47 there have only been 2 studies published to date evaluating the association of circumcision with male-to-male HIV acquisition.48,49 One, a multivariate analysis, found that lack of circumcision was independently associated with a 2-fold increased risk of prevalent infection,49 whereas the other found no association between circumcision and recent infection (but without controlling for behavioral risk).48 Our study found a doubling of the risk of HIV acquisition associated with lack of circumcision, although the PAR in our population was relatively low. There is substantial biologic plausibility supporting our finding of decreased risk of HIV acquisition in circumcised MSM, despite our inability to identify the mechanism by which this occurred. Foreskin mucosa contains an abundance of CD4+ T cells and Langerhans cells, and these cells are further increased in men with recent sexually transmitted diseases.50 Foreskin epithelium from uncircumcised men is more susceptible to infection with CXCR5 viruses than cervical mucosa or keratinized foreskin tissue from circumcised men. Intact foreskin has also been associated with an increased incidence of ulcerative sexually transmitted diseases,51 and recent studies demonstrate that herpes simplex virus 2 (HSV-2) is associated with an increased risk of HIV acquisition among MSM.52 The advisability of promoting circumcision among high-risk adult men is uncertain, because circumcision after puberty may be less protective.53 A randomized controlled trial of this prevention strategy is being planned for heterosexual men in Africa, where the PAR for HIV infection arising from lack of circumcision appears to be substantially higher than in MSM.
Many studies have demonstrated that receptive anal sex is most strongly associated with prevalent12,13,17,54,55 and incident HIV infection in MSM16,18,20 and carries the highest per-contact risk of acquiring HIV.24 This study found that that URA with either HIV-positive or unknown serostatus partners explained one quarter of new infections in this cohort. Surprisingly, we also found an independent increase in the risk of HIV seroconversion among men reporting PRA with an HIV-positive partner. This finding likely represents a combination of overreporting of condom use and unrecognized condom failure by the receptive partner. Condom failure rates are particularly increased among MSM who use condoms infrequently, substance users, and those failing to use appropriate lubricants.56 Public health messages should emphasize the risk associated with PRA as well as URA so that individuals can factor this into their sexual decision making and provide lubricants; such messages should also provide education about proper condom use so as to minimize condom failure.
Although there have been a number of well-documented case reports of HIV acquisition likely as a result of receptive oral sex57,58 and a number of studies demonstrating an epidemiologic association of receptive oral sex with HIV infection,15,16 most studies have failed to find an independent contribution of oral sex to HIV seroconversion after controlling for receptive anal sex, which carries a much higher per-contact risk. Our study found a substantial elevated risk independently associated with receptive oral sex with ejaculation with positive partners, however, after controlling for receptive and insertive anal sex practices. The plausibility of oral transmission of HIV comes from animal studies59 and studies indicating that tonsillar tissue is rich in dendritic cells and M cells, both capable of antigen transport to lymphoid tissue in the absence of trauma or inflammation.60 A longitudinal study of MSM found the per-contact risk of receptive oral sex to be comparable to that of insertive anal sex,24 and several studies of newly infected persons document a substantial minority of newly infected persons reporting only this risk behavior despite repeated questioning.61-63 The fact that other types of contact were reported in our study by seroconverters who also reported receptive oral sex with an HIV-positive partner makes it less likely that our finding is a result of underreporting of stigmatized higher risk behaviors; furthermore, the finding persisted in alternative models in which numbers of contacts were taken into account. Our study was also able to address the independent contribution of unprotected oral sex among MSM engaging in multiple types of risk; studies of persons whose only exposure is to oral sex may not be generalizable to populations engaging in both anal and oral sex. Nevertheless, it is impossible to determine definitively whether the independent association of oral sex with ejaculation in our study represents true transmission by this route or instead may be a marker of riskier sexual practices in general or of unmeasured confounders. Other studies attempting to evaluate the role of oral sex in transmission are difficult to interpret because of the limited number of persons with exposure to known HIV-positive partners64 or because of a focus on serodiscordant couples,65 a group in which the risk of transmission by any route is already likely to be lower than in persons newly entering relationships or persons with multiple partners. Data from our study and reports of newly infected individuals suggest that approximately 5% to 10% of new HIV infections in MSM are attributable to receptive oral sex. Public health efforts should focus on reducing the total number of partners and receptive anal sex contacts; however, individual counseling should include messages about the potential for HIV acquisition to occur by means of oral exposure to infected semen.
High infection rates in MSM populations argue for the need to develop and test prevention strategies for these populations. Early studies were limited in their ability to identify independent risk factors preceding HIV infection. The current study extends earlier work and points to a number of opportunities for building effective prevention strategies: reductions in the number of sexual partners and episodes of receptive anal sex; regular HIV antibody testing and disclosure of test results to all sex partners; and strategies focused on reducing sexual risk associated with substances, including inhaled nitrites.
The authors are grateful for the dedication and commitment of VPS participants, the efforts of VPS site staff, and the contributions of the HIVNET Community Advisory Board Members.
1. Quan V, Steketee R, Valleroy L, et al. HIV incidence in the United States 1978-1999. J Acquir Immune Defic Syndr
2. Weinstock H, Dale M, Gwinn M, et al. HIV seroincidence among patients at clinics for sexually transmitted diseases in nine cities in the United States. J Acquir Immune Defic Syndr
3. Ekstrand ML, Stall RD, Paul JP, et al. Gay men report high rates of unprotected anal sex with partners of unknown or discordant HIV status. AIDS
4. Centers for Disease Control. Advancing HIV prevention: new strategies for a changing epidemic-United States, 2003. MMWR Morb Mortal Wkly Rep
5. Centers for Disease Control. HIV/AIDS among racial/ethnic minority men who have sex with men-United States, 1989-1998. MMWR Morb Mortal Wkly Rep
6. Katz MH, Schwarcz SK, Kellogg TA, et al. Impact of highly active anti-retroviral treatment on HIV seroincidence among men who have sex with men: San Francisco. Am J Public Health
7. Centers for Disease Control. HIV incidence among young men who have sex with men-seven US cities, 1994-2000. MMWR Morb Mortal Wkly Rep
8. Dukers NH, Goudsmit J, de Wit JB, et al. Sexual risk behaviour relates to the virological and immunological improvements during highly active antiretroviral therapy in HIV-1 infection. AIDS
9. Ostrow DE, Fox KJ, Chmiel JS, et al. Attitudes towards highly active antiretroviral therapy are associated with sexual risk taking among HIV-infected and uninfected homosexual men. AIDS
10. Vanable PA, Ostrow DG, McKirnan DJ. Viral load and HIV treatment attitudes as correlates of sexual risk behavior
among HIV-positive gay men. J Psychosom Res
11. Jaffe HW, Choi K, Thomas PA, et al. National case-control study of Kaposi's sarcoma and Pneumocystis carinii pneumonia in homosexual men: part 1. Epidemiologic results. Ann Intern Med
12. Coates RA, Calzavara LM, Read SE, et al. Risk factors for HIV infection in male sexual contacts of men with AIDS or an AIDS-related condition. Am J Epidemiol
13. van Griensven GJ, Tielman RA, Goudsmit J, et al. Risk factors and prevalence of HIV antibodies in homosexual men in the Netherlands. Am J Epidemiol
14. Chmiel JS, Detels R, Kaslow RA, et al. Factors associated with prevalent human immunodeficiency virus (HIV) infection in the Multicenter AIDS Cohort Study. Am J Epidemiol
15. Samuel MC, Hessol N, Shiboski S, et al. Factors associated with human immunodeficiency virus seroconversion in homosexual men in 3 San Francisco cohort studies, 1984-1989. J Acquir Immune Defic Syndr Hum Retrovirol
16. Ostrow DG, DiFranceisco WJ, Chmiel JS, et al. A case-control study of human immunodeficiency virus type 1 seroconversion and risk-related behaviors in the Chicago MACS/CCS Cohort, 1984-1992. Multicenter AIDS Cohort Study. Coping and Change Study. Am J Epidemiol
17. Seage GRR, Mayer KH, Horsburgh CRJ, et al. The relation between nitrite inhalants, unprotected receptive anal intercourse, and the risk of human immunodeficiency virus infection. Am J Epidemiol
18. Schechter MT, Boyko WJ, Douglas B, et al. The Vancouver Lymphadenopathy-AIDS Study: 6. HIV seroconversion in a cohort of homosexual men. CMAJ
19. Kingsley LA, Rinaldo CRJ, Lyter DW, et al. Sexual transmission efficiency of hepatitis B virus and human immunodeficiency virus among homosexual men. JAMA
20. Buchbinder SP, Douglas JMJ, McKirnan DJ, et al. Feasibility of human immunodeficiency virus vaccine trials in homosexual men in the United States: risk behavior
, seroincidence, and willingness to participate. J Infect Dis
21. Koblin BA, Taylor PE, Avrett S, et al. The feasibility of HIV-1 vaccine efficacy trials among gay/bisexual men in New York City: Project ACHIEVE. AIDS Community Health Initiative Enroute to the Vaccine Effort. AIDS
22. Seage GR III, Holte SE, Metzger D, et al. Are US populations appropriate for trials of human immunodeficiency virus vaccine? The HIVNET Vaccine Preparedness Study. Am J Epidemiol
23. Kooperberg C, Petitti DB. Using logistic regression to estimate the adjusted attributable risk of low birthweight in an unmatched case-control study. Epidemiology
24. Vittinghoff E, Douglas J, Judson F, et al. Per-contact risk of human immunodeficiency virus transmission between male sexual partners. Am J Epidemiol
25. Garnett GP, Anderson RM. Strategies for limiting the spread of HIV in developing countries: conclusions based on studies of the transmission dynamics of the virus. J Acquir Immune Defic Syndr Hum Retrovirol
26. Kremer M, Morcom C. The effect of changing sexual activity on HIV prevalence. Math Biosci
27. Anderson RM, Gupta S, Ng W. The significance of sexual partner contact networks for the transmission dynamics of HIV. J Acquir Immune Defic Syndr Hum Retrovirol
28. Fleming P, Byers RH, Sweeny PA, et al. HIV prevalence in the United States, 2000 [abstract]. Presented at the 9th Conference on Retroviruses and Opportunistic Infections, Seattle, 2002.
29. MacKellar D, Valleroy LA, Secura GM, et al. Unrecognized HIV infection, risk behaviors, and misperceptions of risk among young men who have sex with men-6 United States Cities, 1994-2000 [abstract]. Presented at the XIV International AIDS Conference, Barcelona, 2002.
30. Ciccarone DH, Kanouse DE, Collins RL, et al. Sex without disclosure of positive HIV serostatus in a US probability sample of persons receiving medical care for HIV infection. Am J Public Health
31. Woody GE, Donnell D, Seage GR, et al. Non-injection substance use correlates with risky sex among men having sex with men: data from HIV-NET. Drug Alcohol Depend
32. Ostrow DG, Beltran ED, Joseph JG, et al. Recreational drugs and sexual behavior
in the Chicago MACS/CCS cohort of homosexually active men. Chicago Multicenter AIDS Cohort Study (MACS)/Coping and Change Study. J Subst Abuse
33. Purcell DW, Parsons JT, Halkitis PN, et al. Substance use and sexual transmission risk behavior
of HIV-positive men who have sex with men. J Subst Abuse
34. Koblin BA, Chesney MA, Husnik MJ, et al. High-risk behaviors among men who have sex with men in 6 US cities: baseline data from the EXPLORE Study. Am J Public Health
35. Chesney MA, Barrett DC, Stall R. Histories of substance use and risk behavior
: precursors to HIV seroconversion in homosexual men. Am J Public Health
36. Casper C, Wald A, Pauk J, et al. Correlates of prevalent and incident Kaposi's sarcoma-associated herpesvirus infection in men who have sex with men. J Infect Dis
37. Everett GM. In: Gessa GL, ed. Sexual Behavior: Pharmacology and Biochemistry
. New York: Raven Press; 1975:97.
38. Soderberg LS. Immunomodulation by nitrite inhalants may predispose abusers to AIDS and Kaposi's sarcoma. J Neuroimmunol
39. Dax EM, Adler WH, Nagel JE, et al. Amyl nitrite alters human in vitro immune function. Immunopharmacol Immunotoxicol
40. Stall R, Paul JP, Greenwood G, et al. Alcohol use, drug use and alcohol-related problems among men who have sex with men: the Urban Men's Health Study. Addiction
41. Richters J, Gerofi J, Donovan B. Why do condoms break or slip off in use? An exploratory study. Int J STD AIDS
42. Torian LV, Makki HA, Menzies IB, et al. HIV infection in men who have sex with men, New York City Department of Health sexually transmitted disease clinics, 1990-1999: a decade of serosurveillance finds that racial disparities and associations between HIV and gonorrhea persist. Sex Transm Dis
43. Koblin BA, Torian LV, Guilin V, et al. High prevalence of HIV infection among young men who have sex with men in New York City. AIDS
44. Valleroy LA, MacKellar DA, Karon JM, et al. HIV prevalence and associated risks in young men who have sex with men. Young Men's Survey Study Group. JAMA
45. Schwarcz SK, Kellogg TA, McFarland W, et al. Characterization of sexually transmitted disease clinic patients with recent human immunodeficiency virus infection. J Infect Dis
46. Weiss HA, Quigley MA, Hayes RJ. Male circumcision and risk of HIV infection in sub-Saharan Africa: a systematic review and meta-analysis. AIDS
47. O'Farrell N, Egger M. Circumcision in men and the prevention of HIV infection: a “meta-analysis” revisited. Int J STD AIDS
48. Grulich AE, Hendry O, Clark E, et al. Circumcision and male-to-male sexual transmission of HIV. AIDS
49. Kreiss JK, Hopkins SG. The association between circumcision status and human immunodeficiency virus infection among homosexual men. J Infect Dis
50. Patterson BK, Landay A, Siegel JN, et al. Susceptibility to human immunodeficiency virus-1 infection of human foreskin and cervical tissue grown in explant culture. Am J Pathol
51. Moses S, Bailey RC, Ronald AR. Male circumcision: assessment of health benefits and risks. Sex Transm Infect
52. Renzi C, Douglas JM, Jr, Foster M, et al. Herpes simplex virus type 2 infection as a risk factor for human immunodeficiency virus acquisition in men who have sex with men. J Infect Dis
53. Gray RH, Kiwanuka N, Quinn TC, et al. Male circumcision and HIV acquisition and transmission: cohort studies in Rakai, Uganda. Rakai Project Team. AIDS
54. Darrow WW, Echenberg DF, Jaffe HW, et al. Risk factors for human immunodeficiency virus (HIV) infection in homosexual men. Am J Public Health
55. Melbye M, Biggar RJ, Ebbesen P, et al. Seroepidemiology of HTLV-III antibody in Danish homosexual men: prevalence, transmission, and disease outcome. BMJ
56. Stone E, Heagerty P, Vittinghoff E, et al. Correlates of condom failure in a sexually active cohort of men who have sex with men. J Acquir Immune Defic Syndr
57. Lifson AR, O'Malley PM, Hessol NA, et al. HIV seroconversion in two homosexual men after receptive oral intercourse with ejaculation: implications for counseling concerning safe sexual practices. Am J Public Health
58. Lane HC, Holmberg SD, Jaffe HW. HIV seroconversion and oral intercourse. Am J Public Health
59. Baba TW, Trichel AM, An L, et al. Infection and AIDS in adult macaques after nontraumatic oral exposure to cell-free SIV. Science
60. Stahl-Hennig C, Steinman RM, Tenner-Racz K, et al. Rapid infection of oral mucosal-associated lymphoid tissue with simian immunodeficiency virus. Science
61. Schacker T, Collier AC, Hughes J, et al. Clinical and epidemiologic features of primary HIV infection. Ann Intern Med
62. Celum C, Buchbinder S, Donnell D, et al. Early human immunodeficiency virus (HIV) infection in the HIV Network for Prevention Trials vaccine preparedness cohort: risk behaviors, symptoms, and early plasma and genital tract virus load. J Infect Dis
63. Hecht FM, Busch MP, Rawal B, et al. Use of laboratory tests and clinical symptoms for identification of primary HIV infection. AIDS
64. Page-Shafer K, Shiboski CH, Osmond DH, et al. Risk of HIV infection attributable to oral sex among men who have sex with men and in the population of men who have sex with men. AIDS
65. del Romero J, Marincovich B, Castilla J, et al. Evaluating the risk of HIV transmission through unprotected orogenital sex. AIDS
The following institutions and persons associated with the HIVNET participated in the VPS Team: Domestic Master Contractor, Abt Associates (G. Seage and M. Gross); Statistical and Clinical Coordinating Center, Fred Hutchinson Cancer Research Center and University of Washington (T. Fleming and S. Self); Central Laboratory, Viral and Rickettsial Disease Laboratory, California Department of Health Services (H. Sheppard and M. Ascher); Repository Contractor, Biomedical Research (J. Leff); Denver Department of Public Health (F. N. Judson); Fenway Community Health Center (K. Mayer); Howard Brown Health Center (D. McKirnan); New York Blood Center (C. Stevens and B. Koblin); New York University School of Medicine (M. Marmor and S. Titus); Beth Israel Medical Center, New York (D. Des Jarlais); San Francisco Department of Public Health (S. Buchbinder); University of Pennsylvania and the Philadelphia Veterans Affairs Medical Center (D. Metzger and G. Woody); University of Washington (C. Celum); and National Institute of Allergy and Infectious Diseases (R. Hoff, M. McCauley, and Z. Rosenberg).