The hypothesis that the Internet can facilitate HIV serosorting has strong appeal. Individuals may find it easier to disclose their HIV serostatus on profiles or in semi-anonymous chat rooms compared with face-to-face encounters. Moreover, HIV-positive and HIV-negative individuals can simply search profiles for their own serostatus without multiple disclosure discussions. The net effect could therefore be a reduction in HIV transmission.
The Internet may also make it easier to have more sex partners, of seroconcordant or serodiscordant HIV status. The Internet has become one of the most popular venues for men who have sex with men (MSM) to find sex partners and has also been associated with unprotected sex and sexually transmitted disease outbreaks [1–7]. Without certainty of one's own and one's partner's HIV serostatus, the net effect may be to increase HIV transmission if the Internet leads to more unprotected sex with potentially serodiscordant partners. This may be more true for HIV-negative than HIV-positive individuals given that negative serostatus carries more uncertainty and depends upon risks occurring since the most recent test.
To examine the impact of the Internet on HIV serosorting and transmission risk, we conducted a secondary analysis of data from the US National HIV Behavioral Surveillance (NHBS) for MSM. The NHBS was conducted in 2004 in San Francisco (N = 1574) as part of the nationally coordinated system. The methods and primary results of the NHBS on a national level have been described in detail previously . Briefly, the survey is based on randomized, time-location sampling at venues frequented by MSM . The national survey collected demographic information, self-reported HIV serostatus, and sexual risk behavior on an individual level. Our local questionnaire in San Francisco also collected partner-by-partner information, including partner HIV serostatus, sexual behavior within each partnership, and where each partner was first met. We examined 2547 partnerships of 1211 men who reported being HIV negative and 630 partnerships of 251 men who reported being HIV positive. The remaining 112 respondents did not know their own HIV serostatus or had no partner in the past 6 months and therefore could not be used to examine seroconcordancy or unprotected anal intercourse (UAI) in partnerships. Of note is the fact that MSM ineligible for our analysis as a result of the above factors were younger [odds ratio (OR) 0.97 per year, 95% confidence interval (CI) 0.95–0.99] and were less likely to have postgraduate education (OR 0.31, 95% CI 0.14–0.67) compared with those included.
We tested two hypotheses concerning the Internet and HIV risk. First, we examined whether the Internet is associated with an increase or decrease in the likelihood of forming HIV-seroconcordant partnerships (i.e. serosorting) compared with bars or dance clubs. Second, we tested whether the Internet is associated with an increase or decrease in the likelihood of UAI with potentially serodiscordant partners (i.e. partners who are opposite or unknown serostatus) compared with bars or dance clubs. Bars and dance clubs were chosen for the comparison by virtue of being the most common type of venue where partners were met, accounting for 36% of partnerships. The Internet was the second most common venue for meeting partners, accounting for 18% of partnerships. Of note is the fact that few MSM formed all their partnerships in the same way. For example, of men with any Internet-met partner, only 5.7% had exclusively Internet partners in the past 6 months. Of men with any bar or club partner, only 3.5% had exclusively bar or club-met partners. As a result of differences in the direction of associations, data from HIV-negative and HIV-positive MSM were analysed separately. Analyses were performed using generalized estimating equation (GEE) commands in SAS (PROC GENMOD, SAS version 9.1.3; SAS Institute, Inc., Cary, North Carolina, USA) to adjust for the effect of repeated observations on individuals (i.e. multiple partnerships per respondent). Final models also adjusted for the potential confounding effects of age, education, and race/ethnicity (factors associated with Internet use and HIV serostatus).
We found no evidence that the Internet was associated with an increased likelihood of forming HIV-seroconcordant partnerships among HIV-negative MSM compared with bars or dance clubs (Table 1). Among HIV-negative MSM, the direction of the effect was actually towards a decreased likelihood of Internet-met partners being seroconcordant (adjusted OR 0.9, 95% CI 0.7–1.1), although this effect was not statistically significant (P = 0.27). Among HIV-positive MSM, the direction of the effect was towards an increased likelihood of Internet-met partners being seroconcondant (AOR 1.6, 95% CI 0.9–2.7), but was of borderline significance (P = 0.08). The Internet was, however, significantly associated with an increased likelihood of UAI with potentially serodiscordant partners among HIV-negative MSM (adjusted OR 1.5, 95% CI 1.1–2.0, P = 0.01). Among HIV-positive MSM, the Internet was not associated with potentially serodiscordant UAI (adjusted OR 0.9, 95% CI 0.5–1.8, P = 0.75).
Against a background of the increasing use of the Internet to find sexual partners, our data suggest that the net effect of the Internet is to increase the risk of HIV transmission given no significant impact on serosorting and more serodiscordant UAI among HIV-negative MSM. Why there is an increased likelihood of potentially serodiscordant UAI with Internet partners of HIV-negative MSM is not clear. Perhaps initial disclosures or assumptions of HIV-negative status made online were later found in person to be false or uncertain but neither party wished to break the assignation after having gone through the trouble of meeting in person. For example, discussion of risk behavior since the last HIV test or the omission of HIV-positive status on a profile was corrected only after meeting face to face. This situation may happen less often when individuals first meet face to face, then discuss their serostatus, and then decide whether to have sex. Although we do not wish to overinterpret the finding because of its borderline statistical significance in our data, the Internet may be able to provide some advantage for HIV-positive MSM to find seroconcordant partners . Websites can facilitate HIV-positive serosorting through ease of clicking to positive-seeking-positive areas, requirements to disclose serostatus on profiles, and ensuring confidentiality. Because uncertainty will always remain around HIV-negative serostatus, it will always be more difficult to enhance HIV-negative serosorting, online or off.
Conflicts of interest: None.
1. Rietmeijer CA, Bull SS, McFarlane M, Patnaik JL, Douglas JM. Risks and benefits of the Internet for populations at risk for sexually transmitted infections (STIs): results of an STI clinic survey. Sex Transm Dis 2003; 30:15–19.
2. Liau A, Millett G, Marks G. Meta-analytic examination of online sex-seeking and sexual risk behavior among men who have sex with men. Sex Transm Dis 2006; 33:576–584.
3. Elford J, Bolding G, Sherr L. Seeking sex on the Internet and sexual risk behaviour among gay men using London gyms. AIDS 2001; 15:1409–1415.
4. Klausner JD, Kent CK, Wong W, McCright J, Katz MH. The public health response to epidemic syphilis, San Francisco, 1999–2004. Sex Transm Dis 2005; 32:S11–S18.
5. Klausner JD, Wolf W, Fischer-Ponce L, Zolt I, Katz MH. Tracing a syphilis outbreak through cyberspace. JAMA 2000; 284:447–449.
6. Hospers HJ, Kok G, Harterink P, de Zwart O. A new meeting place: chatting on the Internet, e-dating and sexual risk behaviour among Dutch men who have sex with men. AIDS 2005; 19:1097–1101.
7. Bolding G, Davis M, Hart G, et al
. Gay men who look for sex on the Internet: is there more HIV/STI risk with online partners? AIDS 2005; 19:961–968.
8. Centers for Disease Control and Prevention. HIV prevalence, unrecognised infection, and HIV testing among men who have sex with men (MSM) – five U.S. cities, June 2004–April 2005. MMWR Morb Mortal Wkly Rep
9. MacKellar D, Valleroy L, Karon J, Lemp G, Janssen R. The Young Men's Survey: methods for estimating HIV seroprevalence and risk factors among young men who have sex with men. Public Health Rep 1996; 111(Suppl 1):138–144.
10. Davis M, Holt G, Bolding G, Sherr L, Elford J. Sex and the Internet: gay men, risk reduction and serostatus. Cult Health Sex 2006; 8:161–174.