Secondary Logo

Journal Logo

Epidemiology and Social Science

Group Sex Events and HIV/STI Risk in an Urban Network

Friedman, Samuel R PhD*; Bolyard, Melissa PhD; Khan, Maria PhD*‡; Maslow, Carey PhD§; Sandoval, Milagros BA*; Mateu-Gelabert, Pedro PhD*; Krauss, Beatrice PhD; Aral, Sevgi O PhD

Author Information
JAIDS Journal of Acquired Immune Deficiency Syndromes: December 1st, 2008 - Volume 49 - Issue 4 - p 440-446
doi: 10.1097/QAI.0b013e3181893f31
  • Free



Although group sex events (GSEs) among men who have sex with men (MSM) in gay sex venues have been a public health concern since the early 1980s, much less attention has been paid to GSE among other populations. Most public health research on group sex activity among non-MSM1-6 is targeted at early or middle adolescents,7-10 limited to asking 1 or 2 questions about group sex participation among sexually transmitted infection (STI), and family planning clinic patients11,12 or involves ethnographic information from relatively small samples.10,13

GSE range from large events at semipublicly advertized locations to spontaneous events among small groups of friends or acquaintances. They potentially carry great epidemiologic significance for the community's STI/HIV infection rate and the epidemic levels reached. GSEs where high rates of sexual partnership exchange occur may catalyze STI/HIV transmission. In particular, introduction of primary (acute) HIV infection into the sexual partnership pool at a GSE may result in efficient transmission to multiple individuals within a short period of time, given the heightened transmission probability at this stage of HIV infection.14-18 GSEs also may amplify population-level STI/HIV transmission rates if GSE participants have short sexual/injection network paths to large numbers of other people. Coinfection with other STI pathogens19-23 and high-risk sexual partnership patterns-including concurrent partnerships or rapid partner change24-30-may further increase HIV transmission within sexual networks of GSE attendees.16

We used data collected during an HIV risk network study in an impoverished neighborhood in Brooklyn, NY, affected by high levels of STI/HIV to describe the prevalence and characteristics of GSE and to assess potential for STI/HIV transmission risk at GSE.31 We thus describe the prevalence of and respondent factors associated with GSE attendance and participation; GSE-level characteristics; STI and HIV discordance among GSE attendees; and the graphical distribution of GSE attendees within their sexual and drug use networks.



During a sociometric study of risk network patterns of young adults, injection drug users (IDUs), and other populations in Bushwick, 465 respondents 18 years of age or older were recruited, between 2002 and 2004. Bushwick is a primarily Latino section of Brooklyn, NY. Because overall study aims included interest in sexual linkage distances between young adults and IDUs in the community, index cases (“seeds”) were recruited in 1 of several ways: 66 from a population-representative sample of 18- to 30-year-old Bushwick youth recruited door-to-door within randomly selected face blocks; a convenience sample of 38 IDU seeds who had injected drugs within the prior 3 months and either had visible track marks and/or provided other evidence, during detailed verbal questioning, of having injected during the prior 3 months; and 8 seeds who were recruited as participants in a gay sex party subculture. All seeds resided in Bushwick. We also recruited 353 respondents who were risk partners of 1 or more of the 112 seeds or were risk partners of such partners or of their partners. Risk partners were defined as sexual partners, persons with whom respondents attended a GSE-but did not necessarily have sex, or persons with whom they injected drugs (but did not necessarily share syringes/equipment) in the last 3 months. Each index respondent was asked to name and, later in the interview, to provide locator information for up to 10 people with whom they had had sex in the last 3 months; up to 2 with whom they had attended a GSE during the same time period; and, if the index respondent injected drugs, up to 5 people with whom they had injected drugs in the last 3 months. (When we started recruiting attendees at gay group sex parties and their networks late in the project, the maximum number of group sex nominees was increased to 8.) Network sampling consisted of recruiting named partners. These partners were interviewed as well, and we attempted to interview their partners and their partners' partners; thus, 3 “generations” of contacts were recruited. Minor adjustments to these rules are described elsewhere.31

This sample is thus not a probability sample. It shares 1 characteristic with respondent-driven samples and with most other community risk network samples-people with more partners are more likely to be selected. Because recruitment chains were short, and the 35 IDU seeds were not a probability sample, it was not possible to adjust the data statistically for this bias.

Ethical approval for all procedures was obtained by the Institutional Review Board of the National Development and Research Institutes, Inc.


Face-to-face structured interviews were conducted in confidential settings after obtaining informed consent. The interview contained sections on sociodemographics and sexual and drug behaviors and whom respondents had had sex with, attended a GSE with, or injected drug within the last 3 months. In addition, respondents reported whether they had attended a GSE in the past 12 months. Those who had attended a GSE were asked a limited number of questions about their sexual partnerships and condom use during sex at the last GSE they attended. Unprotected sexual activity at GSEs was defined as reporting any sex without a condom at the last GSE attended.

Characteristics of GSEs were also obtained. Based on results from preliminary fieldwork, we asked about 3 types of GSEs: a party with a back room, where some party attendees may go to have sex with 1 or more persons; a threesome, foursome, or larger gathering in which participants get together for the express purpose of having sex; and a party with a professional sex worker, where 1 or more people are paid to have sex with the guests. At each GSE type, some or all attendees engaged in sexual activity. The 3 GSE types were not mutually exclusive, thus attendees may have attended more than 1 type of event in the last year. During analysis, respondents were categorized into 3 mutually exclusive groups based on which of these 3 types of GSE they had most “recently” attended. In this analysis, “other GSE” was defined as parties with professional sex workers or parties where multiple events occurred (eg, parties with both a back room and a professional sex worker).

Respondents were asked to describe how many people were present at the last GSE they attended and the apparent racial/ethnic and gender distributions of these attendees. They were also asked about how many attendees used alcohol and drugs at this event, about sexual partnerships among other attendees, and about condom availability and use at the event.


After obtaining separate informed consent, 10 mL of blood and 10 mL of urine were collected. Blood was tested at Bio-Reference Laboratories using standard methods (Elmwood Park, NJ) for HIV (EIA/WB) and anti-herpes simplex virus (HSV-2) (type-specific FOCUS EIA). Urine was tested for chlamydia (BDProbeTec Amplified DNA assay). See Friedman et al31 for details.

Data Analysis

We used SAS to perform statistical analyses (SAS Institute, Version 9.1.3, Cary, NC). We calculated frequencies of respondent demographic, behavioral, and STI indicators. In bivariable analyses, we examined respondent characteristics by GSE variables including attendance, sexual activity, and unsafe sexual activity at GSEs; respondent sexual behavior at the event by GSE type; and GSE characteristics by GSE type and sex/gender. As mentioned above, statistical adjustment for sampling procedures was not feasible.

Ninety dyads were identified in which both respondents were interviewed and in which at least 1 respondent named the other as someone with whom they attended a GSE in the previous 3 months (though they may not have had sex with each other at this event or, indeed, ever). A dyad was defined as being discordant for a given STI (HIV, HSV-2, or chlamydia) if 1 and only 1 member of the dyad tested positive.

We used UCINET (Analytic Technologies, Version V for Windows, Natick, MA) to construct diagrams of the sexual and drug use risk interaction among study respondents and used these to describe the distribution of GSE attendees within the network structure and the implications of this for community vulnerability.


Sample Characteristics

Greater than half the sample was male (57%) (Table 1). Specifically, the sample was composed of men who have ever had sex with men (MSM) (15%), other men (43%), women who have ever had sex with women (WSW) (18%), and other women (25%). The mean age was 31 years. A majority of the 465 participants was Latino (71%), 20% were blacks, and 10% were of other race/ethnicity.

Sample Characteristics and Group Sex Activities by Select Participant Characteristics

Using a “drug hardness scale” previously described,32 the hardest drug usages reported were IDU (43%), noninjected heroine/cocaine (18%), and crack (15%) (Table 1). Over one quarter of the sample had traded sex for drugs/money in the last year (including 60% of WSW and 59% of MSM). The prevalence of HIV, HSV-2, chlamydia, and any 1 of the above 3 infections was 10%, 49%, 7%, and 58%, respectively. The 10% HIV prevalence reflects the high proportions of IDU and MSM in the sample.

Prevalence of GSE Attendance and Participation

Overall, 36% of the sample attended at least 1 GSE in the 12 months before their interview. Among all participants, substantial minorities had attended a GSE in the last year and participated in sexual partnership (26%) and unsafe sexual partnership (13%) during the most recent GSE attended. Percentages of GSE attendance and sexual activity and unsafe sex during the last GSE were only slightly lower (33%, 22%, and 11%, respectively) among respondents for whom the seed responsible for their being recruited was part of the population-representative youth sample or an IDU and did not differ by which of these samples the seed was from.

Associations Between Individual-level Factors and GSE Attendance and Participation

GSE behaviors were similar across age groups, race/ethnicity, and serostatus on HIV and HSV-2. Attendance, sex, and unsafe sex at GSEs, however, were significantly higher among hard drug users (including IDUs), MSM, WSW, respondents who had traded sex for money or drugs in the last 3 months, gay GSE subculture recruits, and respondents testing positive for chlamydia. Few women other than WSW reported that they had sex at GSE events.

Unsafe sex at a GSE was high among IDUs (82%) and other hard drug users (61%).

Respondent risk behaviors differed by GSE type, with unsafe sex most likely among respondents who last attended a threesome, foursome, or larger sex gathering and least likely among respondents who last attended a party with a back room (Table 2).

Sexual Activity Among Participants at the Last GSE They Attended, by the Type of GSE They Last Attended

Characteristics of GSE

GSE characteristics differed by GSE type (Table 3). The mean number of attendees was greater at parties with a back room (34 persons) and other GSEs (25 persons) than at threesomes, foursomes, or larger sex gatherings (10 persons) (P < 0.001); the respective median values were 30, 3, and 20. Threesomes, foursomes, or larger sex gatherings were relatively intimate groups where most participants engaged in sex; respondents who attended such a GSE reported lower percentages of people they did not know at the event and higher percentages of attendees who engaged in sex, compared with respondents who attended other GSE types. Regardless of GSE type, respondents reported that the majority of attendees at GSEs were high on drugs/alcohol, a minority injected drugs, and the majority engaged in sexual activity: The mean number of sex partners that participants had at the last GSE they attended was 1.5 at parties with a back room, 2.2 at threesomes, foursomes, or larger gatherings, and 2.3 at other GSE events. Condoms were available for use at approximately 70% of GSEs.

Characteristics of Last GSE Attended by GSE Type and by Sex

The characteristics of GSEs attended by men and by women were similar with 1 exception: Compared with women, men attended events where a higher proportion of men had sex with men.

The events attended by participants who had ever injected drugs were similar to those attended by never injectors, with 2 exceptions (data not presented in tables). First, a greater proportion (26%) of attendees at the events attended by ever injectors injected drugs there than at events attended by never injectors (2%; P < 0.0001). In addition, a greater proportion (29%) of women attendees at the events attended by ever injectors had sex with other women there than at events attended by never injectors (17%; P = 0.0114).

Twenty-four respondents who attended GSE were either recruited as attendees at gay group sex parties or were recruited by chain link from them to their partners and beyond. We compared the characteristics of GSEs these 24 participants attended with those attended by the 143 other respondents who attended GSEs (data not presented in tables). They were similar on most variables. However, a greater mean percentage of attendees (88%) had sex at the GSE attended by these 24 respondents than at those attended by the others (76%; P = 0.0075); and a higher percentage of the men at these events engaged in sex with men (48%) than at the events attended by the other 143 (7%; P < 0.0001).

STIs Among GSE Attendees

Sixty-one percent of attendees for whom we have STI data tested positive on at least 1 of 3 infections; 11% tested positive for HIV, 51% HSV-2, and 10% chlamydia (data not shown). Thirty-seven percent of attendees who tested positive on at least 1 of these 3 infections and 34% who tested negative on all 3 infections had unsafe sex at the last event.

STI/HIV Discordance

Among respondents with valid STI results, 12% of 90 dyads were HIV-discordant; (45% HSV-2-discordant and 21% chlamydia-discordant). Too few dyads reported having sex with each other at a given GSE to confidently calculate discordance among sex partners at a GSE.

Attendance and Unsafe Sex at GSEs in the Sociometric Risk Network

Figure 1 graphically displays the locations of respondents who attended GSEs in the sociometric risk network. Participants who engaged in unsafe sex at their last GSE are circled. The large connected component in the center of the figure contains a large number of people who attended such events. Many of the component members who did not attend a GSE reported having had sex with someone who did have sex at such an event, and almost all members of this large component are within a network distance (geodesic distance) of 2 of someone who attended. By way of contrast, many of the smaller components have few or no members who attended a GSE. The clear exceptions are the components in the upper-left of Figure 1, which consist of respondents recruited for their linkages with the gay sex party scene. Unsafe sex at a GSE is reported relatively rarely, although it is reported by a majority of these same upper-left components and also by a cluster of similarly recruited members who seem toward the left side of the large component.

Attended group sex party (pluses) and had unsafe sex at group sex party (circles) by gender/sexuality (MSM = up triangle, WSW = down triangle, other female = circle, other male = square) by hardest drug use ever (from dark red to light pink: IDU, crack, non-injector heroin or cocaine; blue = other) by link type (sex = yellow line, IDU = red, sex, and IDU = blue).


Over one third of respondents in this high-risk sample had attended a GSE in the past 12 months; even among those who were not recruited as part of the gay group sex subculture sample, 33% had attended a GSE in the last year.

GSEs are high-risk environments. There is widespread use of drugs and alcohol at these events, and both sex and unprotected sex are common. Although GSEs vary widely in their number of attendees, many involve 10 or more people having sex at the event. Because respondents averaged more than 1 sex partner at the last event they attended, this suggests that STI transmission to multiple partners at once is possible. Because many respondents reported they did not know many attendees at the GSE they last attended, this suggests the possibility that GSEs may lead to transmission across the boundaries of friendship networks. Thus, GSEs may play an important role in STI/HIV transmission among this high-risk community. These findings point to a need for further research on GSEs among other populations and in population-representative samples. Zule et al13 similarly found that 46% of 41 participants in his ethnographic study of drug users, MSM, and others “knowledgeable about drug use and/or male-to-male sexual activity” in a rural North Carolina county had engaged in group sex.

As was the case in research on gay sex venues, risky sexual practices differed by GSE type.4-6 Threesomes, foursomes, or larger sex gatherings are reported to have the fewest attendees and unknown others and to have the highest percent of attendees reporting condom use-which may give the appearance that this type of gathering is relatively safe. However, this safety is limited: the mean proportion of attendees at these relatively small GSEs who actually have sex is very high (94%), and, within each risk group, participants who attend this kind of event are thus most likely to have unprotected sex despite using condoms for a higher proportion of sex acts if they have sex (data not shown).

The data demonstrate STI discordance among GSE attendees. Substantial percentages of both positive and negative attendees engage in sex and in unprotected sex. As such, there is a serious risk of HIV and other STI transmission at these events.

The potential risk of GSE-induced transmission of HIV or other STIs is not limited to gay men. Diverse groups, including MSM, WSW, other men, a limited number of other women, drug users, and nondrug users, attend GSE, sometimes the same event, and engage in unprotected sex at these events.

As shown in Figure 1, sociometric sexual networks afford considerable opportunity for onward transmission of infections acquired at GSEs. The high proportion of GSE attendees in the large connected component suggests considerable vulnerability to epidemic outbreaks within such networks.

A number of limitations should be noted. First, this sample is a high-risk sample, hence we were unable to measure the prevalence and frequency of GSEs or describe characteristics of GSE attendees among the general population, preventing an assessment of the population-level importance of GSEs to STI/HIV transmission. As is usual in network studies of this type, only a minority of named contacts were reached. In addition, behavioral data are based on respondent recall, not observation, and thus group sex and other behaviors may have been underreported.

As discussed before, GSE are risk situations that bring together heightened levels of behavioral and biomedical HIV risk. At these events, concurrency maximizes behavioral risk, with a possibility that recently acquired HIV infection together with infection by other STIs will amplify biomedical risk. Although population prevalence of GSE is still unknown, the high potential risk and frequency of GSE events in this sample in both non-MSM and MSM participants are worrisome. Researchers should incorporate questions about group sex attendance and sexual behaviors into general sexual behavior surveys and, more generally, in epidemiologic and prevention research. Participants in existing interventions should have the risks of GSEs pointed to them and should be counseled both in terms of avoiding attendance, safety if they should attend, and protecting partners. Interventions should be developed to reduce STI and HIV among attendees at GSEs. Given the lack of experience and research on such interventions, particularly in non-MSM populations, this may require substantial social research.


The authors would like to acknowledge the assistance of the participants in this study.


1. McKusick L, Horstman W, Coates TJ. AIDS and sexual behavior reported by gay men in San Francisco. Am J Public Health. 1985;75:493-496.
2. Clatts MC, Goldsamt LA, Yi H. An emerging HIV risk environment: a preliminary epidemiological profile of an MSM POZ Party in New York City. Sex Transm Infect. 2005;81:373-376.
3. Frankis J, Flowers P. Men who have sex with men (MSM) in public sex environments (Pses): a systematic review of quantitative literature. AIDS Care. 2005;17:273-288.
4. Binson D, Woods WJ, Pollack L, et al. Differential HIV risk in bathhouses and public cruising areas. Am J Public Health. 2001;91:1482-1486.
5. de Wit JB, de Vroome EM, Sandfort TG, et al. Homosexual encounters in different venues. Int J STD AIDS. 1997;8:130-134.
6. Parsons JT, Halkitis PN. Sexual and drug-using practices of HIV-positive men who frequent public and commercial sex environments. AIDS Care. 2002;14:815-826.
7. Dietrich LC. Chicana adolescents: Bitches, 'ho's, and schoolgirls. Westport, CT: Praeger; 1998.
8. Freudenberg N, Roberts L, Richie BE, et al. Coming up in the boogie down: lives of adolescents in the South Bronx. Health Educ Behav. 1999;26:788-805.
9. Krauss B, O'Day J, Godfrey C, et al. Who wins in the status games? Violence, sexual violence and an emerging single standard among adolescent women. Ann NY Acad Sci. 2006;1087:56-73.
10. Rothenberg RB, Sterk C, Toomey KE, et al. Using social network and ethnographic tools to evaluate syphilis transmission. Sex Transm Dis. 1998;25:154-160.
11. Nilsson U, Hellberg D, Shoubnikova M, et al. Sexual behavior risk factors associated with bacterial vaginosis and Chlamydia trachomatis infection. Sex Transm Dis. 1997;24:241-246.
12. Elshibly S, Kallings I, Hellberg D, et al. Sexual risk behaviour in women carriers of Mycoplasma hominis. Br J Obstet Gynaecol. 1996;103:1124-1128.
13. Zule WA, Costenbader E, Coomes CM, et al. Stimulant use and sexual risk behaviors for HIV in rural North Carolina. J Rural Health. 2007;23 (Suppl):73-78.
14. Pilcher CD, Chuan Tien H, Eron JJ Jr, et al, for the Quest Study and the Duke-UNC-Emory Acute HIV Consortium. Brief but efficient: acute HIV infection and the sexual transmission of HIV. J Infect Dis. 2004;189:1785-1792.
15. Jacquez JA, Koopman JS, Simon CP, et al. Role of the primary infection in epidemics of HIV infection in gay cohorts. J Acquir Immune Defic Syndr. 1994;7:1169-1184.
16. Koopman JS, Jacquez JA, Welch GW, et al. The role of early HIV infection in the spread of HIV through populations. J Acquir Immune Defic Syndr Hum Retrovirol. 1997;14:249-258.
17. Wawer MJ, Gray RH, Sewankambo NK, et al. Rates of HIV-1 transmission per coital act, by stage of HIV-1 infection, in Rakai, Uganda. J Infect Dis. 2005;191:1403-1409.
18. Friedman S, Jose B, Neaigus A, et al. Network-related mechanisms may help explain long-term HIV-1-seroprevalence levels that remain high but do not approach population-group saturation. Am J Epidemiol. 2000;152:913-922.
19. 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.
20. Pilcher CD, Price MA, Hoffman IF, et al. Frequent detection of acute primary HIV infection in men in Malawi. AIDS. 2004;18:517-524.
21. Celum CL, Robinson NJ, Cohen MS. Potential effect of HIV type 1 antiretroviral and herpes simplex virus type 2 antiviral therapy on transmission and acquisition of HIV type 1 infection. J Infect Dis. 2005;191(Suppl 1):S107-S114.
22. Freeman EE, Weiss HA, Glynn JR, et al. Herpes simplex virus 2 infection increases HIV acquisition in men and women: systematic review and meta-analysis of longitudinal studies. AIDS. 2006;20:73-83.
23. Wald A, Link K. Risk of human immunodeficiency virus infection in herpes simplex virus type 2-seropositive persons: a meta-analysis. J Infect Dis. 2002;185:45-52.
24. Hudson CP. Concurrent partnerships could cause AIDS epidemics. Int J STD AIDS. 1993;4:249-253.
25. Kraut-Becher JR, Aral SO. Gap length: an important factor in sexually transmitted disease transmission. Sex Transm Dis. 2003;30:221-225.
26. Manhart LE, Aral SO, Holmes KK, et al. Sex partner concurrency: measurement, prevalence, and correlates among urban 18-39-year-olds. Sex Transm Dis. 2002;29:133-143.
27. Morris M, Kretzschmar M. Concurrent partnerships and transmission dynamics in networks. Soc Networks 1995;17:299-318.
28. Morris M, Kretzschmar M. Concurrent partnerships and the spread of HIV. AIDS. 1997;11:641-648.
29. Rosenberg M, Gurvey J, Adler N, et al. Concurrent sex partners and risk for sexually transmitted diseases among adolescents. Sex Transm Dis. 1999;26:208-212.
30. Watts CH, May RM. The influence of concurrent partnerships on the dynamics of HIV/AIDS. Math Biosci. 1992;108:89-104.
31. Friedman SR, Bolyard M, Mateu-Gelabert P, et al. Some data-driven reflections on priorities in AIDS network research. AIDS Behav. 2007;11:641-651.
32. Flom P, Friedman S, Jose B, et al. Peer norms regarding drug use and drug selling among household youth in a low income “drug supermarket” urban neighborhood. Drugs Educ Prev Policy. 2001;8:219-232.

group sex; HIV; sexually transmitted infections; discordant couples; sexual networks; social networks

© 2008 Lippincott Williams & Wilkins, Inc.