Buhi, Eric R. MPH, PhD*; Cook, Robert L. MD, MPH†; Marhefka, Stephanie L. PhD*; Blunt, Heather D. MPH*; Wheldon, Christopher MSPH*; Oberne, Alison B. MA, MPH*; Mullins, Jocelyn C. DVM, MPH†; Dagne, Getachew A. PhD‡
From the *Department of Community and Family Health, College of Public Health, University of South Florida, Tampa, FL; †Department of Epidemiology, College of Public Health and Health Professions, University of Florida, Gainesville, FL; and ‡Department of Epidemiology and Biostatistics, College of Public Health, University of South Florida, Tampa, FL
Supported by the Foundation for the Scientific Study of Sexuality Grants-in-Aid program and a Florida Center for AIDS Research Developmental Core Seed Award.
Correspondence: Eric R. Buhi, MPH, PhD, Department of Community and Family Health, College of Public Health, University of South Florida, 13201 Bruce B. Downs Blvd., MDC 56, Tampa, FL 33612. E-mail: firstname.lastname@example.org.
Received for publication May 12, 2011, and accepted September 1, 2011.
Over the past decade, the Internet has become increasingly important in the romantic and sexual lives of many American adults,1 and a substantial body of research has emerged that explores the sexual health risks associated with online dating and sexual relationships. The majority of these studies have focused specifically on men who have sex with men and suggest that—in this population—meeting online partners is associated with various risk behaviors, including unprotected intercourse,2,3 having multiple sex partners,3,4 substance use,5 a history of sexually transmitted diseases (STDs),5,6 and HIV infection.7,8 Given the growing use of the Internet for dating and sexual relationships among heterosexuals, there is a need to better understand online partnering—particularly among young people, who are at high risk for STDs.9
The objective of this study was to explore sexual health risks among college students based on reports of how they met their last 3 sex partners: online, offline, or some online and some offline. By comparing sexual risk behaviors associated with these 3 divergent types of sex partners, we were able to test 2 competing hypotheses: the “accentuation” hypothesis posits that factors inherent to the Internet increase sexual health risks with partners met online and the “self-selection” hypothesis suggests that use of the Internet as a sexual venue attracts individuals predisposed to risk-taking.10
Data were collected from undergraduates at 2 Florida universities (a large metropolitan university [campus 1] and a large university in a smaller city [campus 2]) during Fall 2008 and Spring 2009. Following IRB approval, a stratified (by year in school) random sample of e-mail addresses was drawn for 6058 (campus 1) and 5481 students (campus 2). Students received an e-mail invitation explaining the study and providing a link to an anonymous, Internet-based questionnaire. E-mail invitations were followed by 3 reminder e-mails, spaced over 30 days.11
A total of 1079 (campus 1) and 1551 (campus 2) students provided data. Respondents were excluded from analysis if they admitted to answering questions dishonestly (n = 27 [campus 1], n = 38 [campus 2]). Another 23 respondents from campus 1 and 4 from campus 2 were excluded due to inconsistent/suspicious response patterns or not meeting inclusion criteria. Data from 2538 students were included in the final dataset, yielding a 22% response rate. Only students reporting as male or female (3 reported being transgendered) and ever having had oral, vaginal, or anal sex were included in the current analyses (N = 2053).
The “partner type” variable was constructed based on respondents' partner-specific reports for each of their last 3 sex partners. Respondents were asked: “How did you first interact with [current/last sex partner]? (First interact means talk with, communicate, etc. for the first time).” Response options were “We first interacted over the Internet/online” or “… in person.” Based on these responses, participants were classified as students with: online partners only; offline partners only; or both online/offline partners. To assess online partnerships beyond the last 3 partners, we asked, “Have you ever had sex with someone offline that you met online?”
Respondents also indicated if they ever willingly had vaginal, oral, and/or anal sex and, if so, age of first sex (for each type) and number of different partners. Risky sex was measured for each of the 3 reported partners with the question, “The first time you had vaginal sex with [current/last sex partner], was a male or female condom used the whole time, from start to finish?” This question was also asked separately for insertive and receptive anal sex. Risky sex for each of the 3 behaviors was defined as a binary variable coded 1 for either “No, only part of the time,” “No, not at all,” or “I don't know,” or 0 for “Yes, from the time we started to when we finished.” Students were asked if they had been tested for HIV in the past year, and if they had ever been told by a doctor or a nurse that they had any 1 of 9 STDs. Unintended pregnancy was assessed by asking: “Have you ever unintentionally become pregnant or gotten someone else pregnant?”
Demographic characteristics and sexual risk behaviors/outcomes were compared between students with online partners only, offline partners only, and both online/offline partners (hereafter called partner type). Odds of demographic and sexual health characteristics were compared between partner types (reference was offline partners only). Mean responses for continuous variables were compared by one-way ANOVAs or Welch test of equality of means. A Tukey post hoc test for pairwise comparisons was used to further elucidate differences between means.
Three generalized estimating equations, using Proc Genmod in SAS 9.2 (Cary, NC), examined the association between meeting a partner online and risky sex (vaginal, insertive anal, and receptive anal) with that partner, while accounting for within-individual correlations. Covariates included categorical age, race, biologic sex, and sexual orientation. Models were run using an exchangeable working correlation structure. Robust empirical standard errors were used in the final model because no appreciable changes in the precision parameter estimates were noted between the exchangeable correlation structure and empirical sandwich estimator.12 Odds ratios with 95% confidence intervals and model-based standard errors were also calculated.
Demographic and sexual characteristics of participants are reported in Table 1. Of the 237 participants reporting sexual activity with a partner met online, 48.8% reported meeting their online partners through social networking sites (SNS), 18.4% through dating or sex-seeking sites, 14.9% through chat rooms, 11.1% through instant messaging, and 6.6% through “special interest” sites (e.g., World of Warcraft). Facebook and MySpace were among the most popular SNS used by participants to meet sexual partners (20.5% and 19.5%, respectively). Fewer participants reported meeting partners on dating sites, such as OkCupid.com (3.1%), Adam4adam.com (2.4%), and Plentyoffish.com (1.4%).
Compared with students with offline partners only, students with online partners only were more likely to be male and gay, lesbian, or bisexual, but less likely to be part of a fraternity/sorority (Table 2). Students with both online/offline partners were older, had lower grade point averages, had a greater number of vaginal and oral sex partners (Table 3), and were more likely to report past-year HIV testing, ever having had an STD, and unintended pregnancy (Table 2). Those with online partners only had a greater number of anal sex partners and were older at first vaginal and oral sex. Sexual intercourse occasions with an online partner were no more likely to involve risky sex compared with occasions with an offline partner after accounting for within-individual correlations (Table 4). This finding was consistent across the 3 sexual behaviors.
Our analyses suggest that meeting partners online does not systematically lead to riskier sex. However, there were notable differences between respondents whose recent sexual partners were met online, offline, or both online and offline. Students who reported meeting partners both online/offline generally had a worse sexual risk profile and experienced more serious sexual health outcomes than students with other partner types.
These data provide mixed support to the accentuation and selection hypotheses. Although it appears that the Internet may facilitate sexual partnerships, risky sex was no more likely with partners met online. Collectively, these findings suggest a more complex relationship between individual characteristics and sexual behaviors than is currently captured by an accentuation versus selection framework. This relationship appears to be somewhat contingent on gender and sexual orientation.
Of note, these hypotheses ignore important contextual factors. For instance, not all students reporting sex with an online partner may have been initially seeking sex. In this study, a higher percentage of students reported having sex with Internet partners (13.5%) than those who reported using the Internet specifically to seek sex partners (9.6%). Furthermore, online sex partners in this study were most commonly met on social as opposed to sexual networking sites, suggesting these encounters did not begin as sexual in nature. These findings are consistent with previous research finding that the majority of online dating encounters resulted in sex, although sex was not always the intention,13 as well as research14 and commentary15 pointing to the complex structure of online and offline social relationships.
This study is limited by self-report cross-sectional behavioral data, a relatively low response rate (22%), potential differences between responders and nonresponders, and data collected from 2 universities. Future research should compare online partnership behaviors of students at universities across the United States.
Since it does not appear that the Internet, itself, is a risk environment, it may be that certain personality traits or other individual differences, such as sexual sensation seeking,16 may characterize young people who meet partners online and increase their sexual risk. Further exploration of such individual differences and collection of biologic STD specimens from students is needed.17
Our findings suggest that online venues might be an important public health intervention context. Prevention efforts might include theoretically-grounded and tailored online risk reduction messages, including condom use promotion and marketing of STD testing sites, appearing as banner advertisements on websites frequently visited by students. Interventions may also include interactive sexual health promotion applications that can be diffused through SNS, as an environment for both STD risk and prevention.18
1.Madden M, Lenhart A. Online Dating. Washington, DC: Pew Internet & Am Life Project, 2006.
2.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.
3.Garofalo R, Herrick A, Mustanski BS, et al. Tip of the Iceberg: young men who have sex with men, the Internet, and HIV risk. Am J Public Health 2007; 97:1113–1117.
4.Mettey A, Crosby R, DiClemente RJ, et al. Associations between internet sex seeking and STI associated risk behaviours among men who have sex with men. Sex Transm Infect 2003; 79:466–468.
5.McKirnan D, Houston E, Tolou-Shams M. Is the Web the culprit? Cognitive escape and Internet sexual risk among gay and bisexual men. AIDS Behav 2007; 11:151–160.
6.Centers for Disease Control and Prevention. Internet use and early syphilis infection among men who have sex with men—San Francisco, California, 1999–2003. Morb Mortal Wkly Rep. 2003;52:1229–1232.
7.Garofalo R, Herrick A, Mustanski B, et al. Online and at-risk: Young men who have sex with men and the Internet. J Adolesc Health 2006; 38:104.
8.Rosser BR, Oakes JM, Horvath KJ, et al. HIV sexual risk behavior by men who use the Internet to seek sex with men: Results of the Men's INTernet Sex Study-II (MINTS-II). AIDS Behav 2009; 13:488–498.
9.Centers for Disease Control and Prevention. Sexually Transmitted Disease Surveillance 2009. Atlanta: GA. Department of Health and Human Services, 2010.
10.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.
11.Dillman DA. Mail and Internet Surveys: The Tailored Design Method. 2nd ed. Hoboken, NJ: John Wiley & Sons, Inc., 2007.
12.Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous outcomes. Biometrics. 1986; 42:121–130.
13.Couch D, Liamputtong P. Online dating and mating: The use of the internet to meet sexual partners. Qual Health Res 2008; 18:268–279.
14.Subrahmanyam K, Greenfield P. Online communication and adolescent relationships. Future Child 2008; 18:119–146.
15.Rietmeijer CA. You've got mail. (2010). Sex Transm Dis 2010; 37:486–487.
16.Horvath KJ, Beadnell B, Bowen AM. Sensation seeking as a moderator of Internet use on sexual risk taking among men who have sex with men. Sex Res Social Policy 2006; 3:77–90.
17.Al-Tayyib AA, McFarlane M, Kachur R, et al. Finding sex partners on the Internet: What is the risk for sexually transmitted infections? Sex Transm Infect 2009; 85:216–220.
18.Rietmeijer CA, McFarlane M. Web 2.0 and beyond: Risks for sexually transmitted infections and opportunities for prevention. Curr Opin Infect Dis 2009; 22:67–71.