The Internet profiles of men who have sex with men within bareback websites

Abstract

The Internet has become a venue for men who have sex with men (MSM) to search for sexual partners. Some of these men intentionally seek unprotected anal intercourse with other men (a.k.a., “bareback” sex). This paper focuses on the creation, use,and content of Internet personal profiles of MSM in the greater New York Citymetropolitan area whouse bareback sites for sexual networking. We used a mixed-methods approach to examine data from a cybercartography of Internet sites conducted during the first phase of the research (199 personal profiles), and from in-depth interviews conducted during its second phase (120 MSM who sought partners online for bareback sex). Results indicate that men generally followed offline stereotypical patterns in their online profiles. However, men who disclosed being HIV positive were more likely to include face and head pictures. Overall, the images they used were heavily sexualised in accordance with group norms perceived and reinforced by the websites’ design and imagery. Bottom-identified men tended to be more explicit in the exposition of their sexual and drug use interests online. This paper highlights how certain virtual and social performances play upon and reinforce other, in the flesh, performances.

Introduction

Personal profiles have emerged as a salient form of self-presentation online and, by extension, offline. They have gained in popularity such that some online sites, namelyfacebook.com (which was created to share personal profiles among college students), has grown exponentially, currently with more than 500 million users worldwide (Wortham 2010). The relational capacities of such sites have not gone unnoticed by individuals seeking others for romantic or sexual partnership, and numerous websites exist for the purpose of facilitating real life encounters, such as match.com, friendster.com, and j-date.com. Similar to personal ads in traditional print media, these websites draw upon idiomatic codes that communicate certain assumptions about what is sexually desirable in a mate (Heino, Gibbs, and Ellison 2006; Walther, Loh, and Granka 2005). The majority of these are features that would probably be immediately apparent or discernible in face-to-face encounters, but must be rendered explicit in online communication. For this reason, online profiles can be a rich source of analysis for researchers who wish to understand the expectations, desires, and social morays of the users of networking websites, who represent themselves and interact with one another in a virtual space, many with the hope of translating their online communications into offline encounters (Carvalheira and Gomes 2003).

Websites used by men who have sex with men (MSM) to find sexual partners are especially relevant sources of information for researchers interested in studying the interplay between sexual desires and norms, on one hand, and sexual behaviour, on the other. For instance, personal ads and profiles placed on Internet websites underline MSM's tendency to emphasize their stereotypical masculine traits (Phua 2002), and also their straightforwardness about their sexual interests (Phua, Hopper, and Vazquez 2002; Tewksbury 2006).

With the rise in incidences of sexually transmitted infections (STIs) among men who have sex with men (Jaffe, Valdiserri, and De Cock 2007), such websites provide insight into sexual practices that may contribute to the spread of STIs, such as intentional unprotected anal intercourse, or “bareback” sex (e.g.,Carballo-Diéguez et al. 2009; Elford et al. 2001). In recent studies, most MSM failed to mention safe sex intention in their online Internet profiles or ads (Blackwell 2009; Drowning 2010). Other studies (Grov 2010; Sowell and Phillips 2010) identified a low number of such ads and profiles mentioning safe sex, although an even lower numbermentioned intentional unsafe sex.

Tewksbury (2003) analysed a sample of profiles posted on a website specialising in bareback sex. His results showed that the majority of barebackers were not specifically seeking HIV infection. Another study by the same author (Tewksbury 2006) focused on profiles of self-proclaimed bug-chasers (men actively pursuing HIV infection) and gift-givers (HIV-positive men willing to infect others), confirming that these are but a minority of all those seeking bareback sex. The majority of men were in their 30's and White. Most profiles were sexually explicit and only 30% included face pictures. Most men indicated no preference concerning their partners’ serostatus. Two-thirds were interested in drug use associated with sexual activity.

Profiles found on websites specialised in bareback sex reveal not only the intentions and desires of their users, but also detail the broader sexual culture in which men who use them participate and “perform.” Drawing from Goffman's (1959) theory of performance, we understand the creation of online profiles as a form of self-presentation whereby MSM choose to highlight and/or exclude certain personal traits in the hope of attracting sexual partners. The end point of this “performance”, the sexual act, is itself another performance, albeit one that lies beyond the observation of social scientists. However, we can try to infer the consequences of that hidden performancefrom its more visible signs: the online profile that anticipates the sexual encounter. Ross (2005) described the Internet as a sexual space midway between fantasy and action that allows sexual experimentation and interaction with the fantasised other. This sexual enactment may well later be transposed into a real life encounter. Studying online profiles may, therefore, give us insight into intentions, decision-making, fantasies, and desires that can inform our intervention strategies with this population.

This is even more relevant when trying to understand sexual practices that may impact individuals’ health status, such as barebacking. To study barebacking without considering the social contexts and spaces that facilitate such behaviour is to miss a critical aspect of the phenomenon. Therefore, this paper describes the online profiles of MSM who used the Internet to find bareback sex partners and the decision process behind their creation, in an effort to understand how certain virtual and social performances play upon and reinforce other, in the flesh, performances.Our research questions were: What are the demographic and behavioral profiles of men who seek male sex partners on bareback websites?Howdo they use virtual images order to meet sexual partners and to portray the intended sexual acts they seek?

Methods

This paper is based on data collected for a study that examined intentional unprotected anal intercourse (“bareback” sex) among men who use the Internet to meet sexual partners. Our analyses focus on quantitative data resulting from a collection of online profiles found on the six most popular websites used by men in the New York City area to meet other men forbareback sex (Phase 1, see [deleted for anonymity] for a description of the cybercartography of Internet sites, amethodology used to select and study websites) and qualitative data from face-to-face interviews of MSM recruited online from those websites (Phase 2, see [deleted for anonymity]).The profiles of the volunteers in Phase 2 may or may not have been selected in Phase 1 of the study; thus, the samples for analysis, despite originating from the same population, are nevertheless different. The Institutional Review Board at the New York State Psychiatric Institute approved this study.

First phase

A systematic method was used to select profiles from the six websites under study. With the goal of selecting five profiles per data collection sessionin one week in June of 2004, we accessed the websites in a random order during 126 randomly-selected hours and selected the n^th profile based in the NYC area from those that were presented as being logged in on the sites or from those that resulted from searches using website engines. For example, if a search of profiles in the NYC area resulted in 100 profiles, we selected every 20^th profile. Duplicates were eliminated, including identical profiles on more than oneof the websites selected for this study. We present results from profiles selected from the two bareback sites only (n=199).

We first obtained frequencies on user characteristics, whether drug use was indicated, and whether a photo was included in the profile. User characteristics included racial/ethnic group (White:n=114, 59%; Latino:40, 21%; African American/Black: 27, 14%;and “Other” ethnicity: 12, 6%), age group (29 or younger: 45, 23%; 30-39: 98, 49%; and 40 or older: 56, 28%), sexual orientation (gay: 174, 88%; and bisexual: 24, 12%), HIV status (HIV-positive: 31, 15%; HIV-negative: 107, 54%; and not provided: 61, 31%), and preferred sexual position (bottom: 35, 18%; versatile bottom: 47, 24%; versatile: 47, 24%; versatile top: 25, 12%; and top: 45, 22%). Among profiles with photographs, we obtained frequencies for 15 dichotomous variables indicating whether the photo had a particular characteristic (see Table 1).

Table 1.

Crosstabulations of user's characteristics, picture inclusion and type of picture

	No. with Pictures (%)a	FULL BODYb	TORSOb	FACE HEADb	PENETRATIONb	ORALb	BUTTb	ERECTIONb	MULTIPLE PICSb	DUO OR GROUPb
Total (N = 199)	134 (67%)	40 (30%)	32 (24%)	67 (50%)	26 (19%)	9 (7%)	71 (53%)	73 (54%)	93 (69%)	29 (22%)
ETHNIC GROUP	NS				X2(3, 130) = 6.52 p = .09c
White (114)	81 (71%)				20 (25%)
Latino (40)	27 (68%)				1 (4%)
African American/Black (27)	14 (52%)				4 (29%)
Other (12)	8 (67%)				1 (8%)
AGE GROUP	NS	X2(2, 134) = 7.84 p = .02		X2(2, 134) = 7.50 p = .02		X2(2, 134) = 4.83 p = .04c
29 or younger (45)	32 (71%)	14 (44%)		22 (69%)		2 (6%)
30-39 (98)	68 (69%)	13 (19%)		27 (40%)		2 (3%)
40+ (56)	34 (61%)	13 (38%)		18 (53%)		5 (15%)
SEXUAL IDENTITY	NS	X2(1, 134) = 4.83 p = .04c	X2(1, 134) = 5.70 p = .01c	X2(1, 134) = 4.54 p = .06c
Gay (174)	118 (68%)	39 (33%)	32 (27%)	63 (53%)
Bisexual (24)	16 (67%)	1 (6%)	0 (0%)	4 (25%)
HIV STATUS	X2(2, 199) = 4.94 p = .09	X2(2, 134) = 6.83 p = .03		X2(2, 134) = 13.04 p < .01				X2(2, 134) = 4.54 p = .10	X2(2, 133) = 6.02 p = .05
Positive (31)	26 (84%)	12 (50%)		20 (77%)				19 (73%)	23 (89%)
Negative (107)	67 (63%)	15 (22%)		34 (51%)				33 (49%)	45 (67%)
Not provided (61)	41 (67%)	12 (29%)		13 (32%)				21 (51%)	25 (64%)
POSITION	NS				X2(4, 134) = 14.63 p = .01		X2(4, 134) = 34.75 p < .01			X2(4, 134) = 9.69 p = .05
Bottom (35)	21 (60%)				10 (47%)		19 (91%)			9 (43%)
Top (45)	31 (69%)				4 (13%)		6 (19%)			6 (16%)
Versatile (47)	35 (75%)				3 (9%)		17 (49%)			4 (11%)
Versatile top (25)	15 (60%)				2 (13%)		5 (33%)			2 (13%)
Versatile bottom (47)	32 (68%)				7 (22%)		24 (75%)			9 (28%)

Note. NS: Non-significant. Non-significant chi-square values were found for: Nudity (n = 121, 90%), Headless (n = 100, 75%), Genitals (n = 107, 80%), Semen (n = 6, 4%), Toy (n = 3, 2%), or Fetish (n = 57, 42%); and therefore, these are not listed in this table.

^aPercentages represent proportions of the total number of profiles for the corresponding group (i.e., 81 out of the 114 profiles of White men included at least one picture).

^bPercentages represent proportions of the total number of profiles containing at least one or more pictures for the corresponding group (i.e., 20 out of the 81 White men's profiles with at least one picture depicted penetration).

^cThe Chi-square statistic is not valid since one cell has an expected count of < 5; therefore, the p-value for Fisher's Exact Test is reported.

In order to determine if there were significant differences in photo content by user characteristics, Chi-square analyses were performed using SPSS 19.0. Based on results examining the groups overall (see Table 1), we partitioned the groups to 2×2 cross-tabulations to facilitate interpretation of between group differences using the criterion of p<.05.

Second phase

The second phase of the study consisted of face-to-face interviews of 120 MSM recruited online from the websites identified in Phase 1 about their sexual practices and Internet use to meet men for bareback sex (see [deleted for anonymity] for a detailed description of the methodology used). Eligible participants were male and at least 18 years old; lived in the NYC metropolitan area; had had intentional condomless anal intercourse at least once with a male partner met over the Internet; and used the Internet at least twice a month to meet sexual partners. A sampling design was developed in order to recruit similar numbers of MSM from different ethnic backgrounds, including Latino, European American, African American, and Asian or Pacific Islander, and tooversample HIV-negative men who reported having had unprotected receptive anal intercourse.Individuals who qualified were scheduled for interviews at our research offices, during which theywere required to undergo an informed consent process before their interview.

Measures

A semi-structured interview guide was developed to cover a variety of topics, including the frequency of Internet use, barebacking, condom use, HIV testing, and non-condom HIV prevention strategies, among other issues. Although the guide had a predetermined structure, it was flexible and the interviewers could follow the natural flow of the information presented by the interviewee.As part of the interview, participants were asked to describe their profiles. To help with recall, participants could review their profiles to see what their profiles looked like and the type of information they had included in them. The qualitative results drawn from this part of the study reflect the interviewees’ descriptions of and reasons for the type of information they included in creating their profiles and reasons for choosing not to include other type of information. The interview lasted approximately two hours, and participants were paid $50 as compensation for their time.

Analysis

The initial coding paralleled the structure of the interview guide to capture the major themes (e.g., one code was “Profile”; see [deleted for anonymity] for a detailed account of the coding process). The sections of the transcripts that covered any reference to profiles including its creation were coded using NVivo 7.0 software for qualitative content analysis (QSR 2008). We analysed the material coded under “Profile” and selected quotes to provide accounts of participants’ choices in creating and managing online profiles.

Results

We opted for a mixed presentation of results, prioritising the quantitative results from Phase 1 (hereafter P1) and using quotes from the qualitative phase (P2) to illustrate the topics covered by the former. We do not identify websites, profile handles, or participant names; instead, we refer to pseudonyms or participant ID numbers.

Profile characteristics

A total of 199 systematically selected profiles were examined. In examining cross-tabulations among demographic variables, several significant Chi-squares were obtained (p<.05). A significantly larger than expected percentage of White men (94%) identified as gay in their profiles compared to African American or Black (69%) and Latino (83%) men combined (χ²_{(1, 199)}=10.72);more Black and Latino men identified as bisexual than White men. A significantly larger percentage of men 40 and older (34%) reported being HIV-positive in their profiles compared to younger men (17%; χ²_{(1, 138)}=5.01). None of the 24 bisexual-identified men indicated a positive HIV status in their profile (four did not provide HIV status and 20 indicated being HIV-negative), while 25% of the gay-identified men indicated that they were HIV-positive (χ²_{(1, 137)}=6.85).

Drug use

Thirty-eight percent of men indicated in their profiles that they used drugs. A larger percentage of White and Latino men (47%) indicated in their profile using drugs compared to other ethnicities (23% of African American/Black men and those who indicated “other” combined; χ²_{(1, 175)}=6.77). Fewer men over the age of 40 indicated drug use (28%) as compared to younger men (47%; χ²_{(1, 181)}=5.55). A marginally significant difference was found with respect to HIV status: more HIV-positive men indicated drug use than those who chose not to provide their HIV status (χ²_{(1, 81)}=3.30; p=.07), but not more than men who indicated being HIV-negative (χ²_{(1, 126)}=0.81; p=.37). Finally, fewer men (29%) who reported being a “top” or “versatile top” in their profiles indicated using drugs as compared to men who noted being “versatile,” “versatile bottom” or “bottom” (50%; χ²_{(1, 181)}=7.43). No differences were found in drug use by sexual identity.

Pictures in profiles

The number and percentages of profiles that included a photograph with a particular characteristic are listed in Table 1. Of the total profiles examined, two-thirds (n=134, 67%) included at least one photograph, and many profiles had multiple pictures. For some of the men, concern that friends would recognise their photos led to the decision to not post a photo (P2):

I didn't include a picture because I created the profile to cruise for guys and if my friends saw me in a chat room, I didn't want to be, like, “A-ha, here you are.”

(Jose, 20 y.o., Latino, HIV-)

[MEAT] is like a secret site. I know none of my friends are on there. So that one, I have -- they have the category bondage, and voyeurs and exhibitionist, all that kind of stuff. I have everything marked, three-ways, groups, bondage, water sports. They will see it inside, but I don't have a picture there. So you have to contact me. And I'll unlock my pictures so that you can see them. And on my good one I have a big picture of me sitting on my deck reading a book.

(Antony, 35 y.o., mixed ethnicity, reported HIV- but never tested)

The second participant, particularly, discussed how he managed his different profiles and his exposure towards his friends, considering the type of profile he used (in this case, a more sexually explicit profile versus a milder, “good-behaviour” profile).

But deciding not to include a picture on a profile may serve another function. For instance, it may be a way to negotiateaccess to other men's photos:

I: Do you put pictures of yourself or not?

R: Not in my profile, I'd rather somebody ask me for it, that way I can get their picture. If I put my picture in my profile, they're not going to want to send me their picture.

(Gabriel, 26 y.o., White, HIV+)

Type of pictures

The 134 profiles containing photographs were examined for specific characteristics (P1). Often men had more than one picture in their profiles, therefore percentages are not cumulative.A larger proportion of HIV-positive men (84%) included a photograph in their profile relative to men who reported being HIV-negative (63%, χ²_{(1, 138)}=4.94). The comparison between HIV-positive men and those who did not provide an HIV status in their profile did not reach statistical significance (67%; χ²_{(1, 92)}=2.88; p=.09).

Several significant findings were obtained in conducting Chi-squares for the cross-tabulations between demographic variables and photo characteristic variables within the subsample of profiles that included photographs. These findings were broken down into the type of photographs displayed.

Body pictures

A significantly greater percentage of men 29 or younger and 40 or older (44% and 38%, respectively) included full body photos in their profiles than men ages 30-39 (19%). More gay-identified men (33%) as compared bisexual men (6%) included a full body photo. Similarly, 27% of gay-identified men who included any sort of picture in their profiles included at least one torso photo in their profile. Finally, a larger proportion (50%) of men who noted being HIV-positive in their profiles included a full body photo as compared to HIV-negative men (22%; χ²_{(1, 93)}=6.79).

Showing different non-sexual parts of the body may encompass different meanings and messages (P2):

I: And tell me how people typically approach you to make contact.

R: Well, first, they like my photo. Because -- Well, I don't have a face photo, but I have very stylized, very artfully, tastefully done body shots that are provocative. And so that's how.

(Jack, 31 y.o., API, HIV-)

It's a picture of me kicking back in a black tank top. And I like this picture because, right after I lost a lot of that weight, and I started working out, and I was just starting to get the self-confidence that I have now, I wanted a decent picture for my profile. I was at a friend's house. And I was just sitting back in a chair, relaxing. And he took my picture. And this is that picture. And I look good in that picture. I thought to myself, “Wow! That's me! Wow!” So I kept it. It's a personal favourite. It's a sentimental favourite.

(Manuel, 35 y.o., Latino, HIV+)

In both quotes it is possible to identify a careful, personalised choice of the body pictures. The images do not seem to be primarily sexually oriented. Yet, they portray a certain intimate dimension of who these men are.

Face pictures

There were several findings with regard to face/head photographs. A greater proportion of men ages 29 and younger included a face/head photo in their profile as compared those 30-39 years of age (χ²_{(1, 100)}=7.35). Also, more gay-identified men included face/head photos in their profile as compared to bisexually-identified men. Lastly, a greater number of HIV-positive men included face/head photos in their profile as compared to HIV-negative men (χ²_{(1, 93)}=5.27) and men who did not specify their HIV status (χ²_{(1, 93)}=13.01).

Some of the men interviewed for P2 explained how they considered Internet-initiated contacts to be basically anonymous, that being the main reason for them not to show their face in their profile pictures.

I don't ever have a face picture. And I may be the only person that you hear that says that. Or maybe I haven't been. But I've been adamant about not having a face picture. To me, I think it's preposterous to have one. Because this is supposed to be a blind meeting, and a lot of the time, when you're meeting people using a face picture, sometimes the person isn't who they say they are.

(Mitch, 25 y.o, Latino, HIV-)

If they say, “Oh I want to see your body, your face.” Uh-uh. And I sometimes even list on the profile, “Don't send face pics so don't ask” so I like to keep the situation very anonymous and very disconnected. So, it's all about a disconnected experience. The whole thing, I like to keep -- I'm very business-like about the whole thing.

(Blake, 32 y.o, African American, HIV+)

For these men, of different HIV status, choosing not to show their faces is a sort of ideological statement. It is not apparently related with privacy concerns, but with an idea of what Internet sexual networking “should” be – basically the first step for anonymous or uninvolved encounters between men. This option thus seems to be rooted in specific beliefs or core values of the men.

There were, however, some men that did mention issues that were closer to privacy concerns for not showing their faces:

In my profile, you will find a link that when you click that link, and this is also for time consumption, instead of asking me for the four photographs, that links you to a site that will have pictures that I've already selected. It won't show my face, but it will show basically everything else. It took me a lot of courage to make those pictures.

(John, 40 y.o, African American, HIV+)

In light of these results, it is interesting to look into the reasons of men who decided to show their faces online.

If [the website is] free, I'll just put a profile up; if someone writes to me, I'll write to them. Otherwise, I won't start writing to people. And then I moved my pictures, 'cause I use the same pictures in all my profiles. And it's all -- there's a face picture with just regular upper-body shots with a jacket on. I mean, I don't have any naked pictures. So I thought whoever sees it, and they see it on another profile, they'll know it's the same person.

(Albert, 27 y.o., API, HIV-)

For this participant using face pictures provides a sense of identity throughout the different profiles he uses. It has a social function, allowing others to identify him whenever they run into his profiles across the web. Other participants discuss similar intentions by using the same screen name in different profiles they created and simultaneously managed.

Sexually explicit pictures

Even more than pictures of naked bodies, images of sexual activity (two or more persons engaging in actions involving genital stimulation) or pictures portraying sexually charged parts of the body such as those of butts or of penises, are the ones that, when included in personal online profiles, more clearly inform about the nature of men's intentions. The following quote (P2) tells of that:

I usually don't include any kind of stats unless they put the age stuff in there. Because I'm just like beyond stats. I'm like, “If you like the pictures, you're going to respond to me. If you don't like the pictures, then don't respond to me.” Pictures are a very big, important part of my profile. (...) But my pictures are very explicit and I usually always show penetration. And that usually always gets them going. Because I know -- I mean, I realise that gay men are visual. And I'm actually a pretty good photographer. So, I mean, it's definitely, you know, hit the level of beyond erotic art and to like, nearly porn. But still, it's very hot. It's like, you look at it, you're like-- oh man!

(Blake, 32y.o, African American, HIV+)

This participant is aware that sexually explicit images are a good way to get attention for his profile, but also a way to filter those who likely will or will not contact him. Thus, it seems that the choice of images posted online is strategic.

There were also several findings regarding photographs depicting penetration and oral sex. A larger proportion of White and African American men (25% and 29%, respectively) included a penetration photo in their profile compared to Latino men (4%; χ²_{(1, 122)}=5.99). Photos depicting penetration were more common in those who indicated “bottom” (48%) than the other position groups (14%; χ²_{(1, 134)}=12.68).

In examining profiles that included oral sex photos, a significant larger proportion of men ages 40 or older (15%) included an oral sex photo in their profile as compared to men ages 30-39 (3%) and those 29 and younger (6%). Here is a quote from one of the interviewees about the oral sex picture in his profile (P2).

So -- and this picture. It's me with somebody's dick in my mouth. I met this guy on [website X]. He seemed like a really nice guy. After we had sex, I realized he was psycho. And his face is not in the picture. [...] The picture was enough -- it'll scare some people away. It'll attract others.

(Manuel, 35 y.o., Latino, HIV+)

For this participant, as for the previous one quoted, the explicit nature of the picture used in his profile seemed to have the purpose of filtering specific types of website users.

More men who identified themselves as “bottoms” or “versatile bottoms” in their profiles included butt photos (91% and 75%, respectively). Fewer (19%) men who identified themselves as “tops” in their profiles included a butt photo. A larger proportion of men who noted that they were HIV-positive in their profile (73%) included a photo of an erect penis as compared to men who indicated being HIV-negative (χ²_{(1, 93)}=4.31).

Multiple pictures and sexual position

Finally, a larger proportion of HIV-positive men (89%) included multiple photos in their profile as compared to HIV-negative men (67%; χ²_{(1, 92)}=4.48) and to those without mention of an HIV status (61%; χ²_{(1, 67)}=5.92). Moreover, more men (34%) who identified themselves as “bottoms” or “versatile bottoms” in their profiles included duo or group photos (43% and 28%, respectively) than men who identified as “tops,” “versatile,” or “versatile tops” (14%; χ²_{(1, 134)}=7.85).

There were no statistically significant findings with regard to nude, headless or genital photos, photos showing semen, involving use of a sex toy, or involving fetish behaviour.

Discussion

The purpose of this study was to describe the online profiles of men who use the Internet to find bareback sex partners, in an effort to gain insight into the motivations of those who create them, within the context of their potential risk behaviour associated with using the Internet to meet other men. We used a mixed methods approach by systematically analysing the content of online profiles from websites used by men to find partners for bareback sex, and by drawing from in-depth interviews to understand some of the meanings that the users of such profiles attribute to them. This strategy allowed us to have a broader and contextual perspective not only of these men's online presentations, but also of the reasoning and decision-making behind some of their choices during the construction and maintenance of theirInternet profiles.

The general findings of the study contribute to the understanding of how men actively use their online images and performances to put themselves in diverse categories, including those of sexual position and identity in order to appeal to potential sexual partners. In addition, the dramatic effect of the performance is reinforced by using stereotypes of certain sexual and ethnic groups, as suggested by Wilson and colleagues(2009). These researchers found that sexual stereotyping is an important mechanism for making sexual partnering decisions and for making sense of the milieu of sexual images presented to the participant within bareback websites. This approach may help men begin to fantasize and desire for their face-to-face sexual encounter, by knowingly carrying out sexual activities based on the images of the online profiles (Ross 2005).

Therefore should be no surprise that the analysed profiles tended to be highly sexualized, which may be a partial consequence of the design of the websites where they were hosted. These include pre-determined fields for detailed information on intended sexual activity, HIV risk reduction strategies, and sexual position in anal intercourse, to name a few. As reported by Carballo-Diéguez and colleagues (2006), the categories and options provided by the web masters of such websites generally determine the ways people portray themselves. We were also able to verify that, when men were free to enter text and photos in open fields, they depicted themselves in sexually explicit ways.

Websites directed at men interested in bareback sex also tend to include explicit imagery with pictures of anal intercourse where no condoms are used, as well as images of naked muscular men with erections (Carballo-Diéguez et al. 2006).Although it was unclear to what extent this imagery influenced men's self-presentation, there seems to be a highly dynamic, eroticized performance on online profiles, in which online self-presentation is limited and influenced by the website structure and design, but it is also actively constructed by the users. Simultaneously, the websites are also shaped by content added by users, on a continuous constructivist process where the website structure and individual self-presentation mutually influence and reinforce each other.

The analysis of explicit sexual performance is particularly interesting in the case of men who, in their profiles, identify as bottoms or versatile bottoms (those who have a preference for being sexually receptive during anal intercourse). Our findings suggest that both, as expected, had significantly more butt pictures in their profiles than any other men, but also more penetration and group pictures. This suggests a more direct exposition of their sexual role and preferences. Furthermore, these men also tended to mention they were more interested in using drugs associated with sex. Therefore, several high-risk practices are suggested in these men's profiles long before any physical action takes place. The theme (bareback sex) is given by the website, the actor's sexual position and interest in drug use is specified by the options that the website allows, and the picturesthat the men post set imagery suggestive of the sexual act. Thus, in several ways, the stage is set up for the in the flesh performance to take place. These results highlight the fact that many high-risk behaviours are anticipated by fantasy, which in turn translates into scenarios created by the men in articulation with the website's possibilities of how they would like situations to unroll and which practices they wish to enrol in, which generally include identifiable elements of risk.

Other authors have indicated that individual preference and mutual agreement are the main reasons for bottom-identified men not to use condoms with partners met online (Ostergren, Rosser, and Horvath 2011). Our findings suggest that online profiles may be the contextual and symbolic link between these two elements in sexual risk taking among men who seek sex partners on online bareback sites.

Drug use deserves special attention, because using substances that alter mood and judgement can lead to higher levels of risk taking, even for men who already choose to bareback. According to research, men who use illicit drugs tend to use Internet-based sex sites to find other users. Thus, the risk factor of illicit drug use is amplified because both sexual partners have found compatibility based on each man's illicit drug use (Blackwell 2008).

HIV status was the category of analysis where more significant differences were found. HIV-positive men were more likely to post more pictures in their profiles, but also to post more body, erection, and face/head pictures. This was a somewhat counterintuitive result, considering the social stigma that still surrounds a positive HIV status in society (Wolitski,Parsons, and Gómez 2006).

All of these results, but this last one in particular, seem to suggest that HIV-positive men who seek sex partners in online bareback sites may feel comfortable including their serostatus in their online presentations, and therefore may not fear HIV- associated stigma or are willing to accept the risk. These are websites where HIV status is, after all, put into perspective, where other categories are prioritised and where risk behaviours are reinforced (Tewksbury 2006). On the other hand, there is evidence that disclosure is one of the strategies used by MSM to cope with discrimination. In a qualitative study about how HIV-positive MSM dealt with stigma,disclosure was identified as one of the leading ways that men affirmedsocial support systems (Chenard 2007). Perhaps, this is a strategy that men are also using online, and that by showing several pictures of themselves, including some of their faces and erect penises, they can assure potential partners that they look healthy and are sexually desirable and capable. By arousing their viewers, the creators of the profiles may be, willingly or not, increasing their chances of attracting sexual partners. However, the shade of social judgement is still present, at least for some men, like the ones quoted saying how they were self-conscious about the possibility of some of their friends discovering they were using bareback websites.

These concerns suggest that for some men a part of their online sexual image and activity remains hidden, as a performance not assumed or discussed out loud. This concurs with the resultsof Carballo-Diéguez and colleagues(2009), where it was found that, despite their sexual practices, some men were uncomfortable with admitting to not using condoms at least sometimes, as they were with the possibility of being considered a “barebacker”. This result, apparently contradictory with the willingness with which some men show their faces, points to the fact that, far from it being a homogenized group of men, those who use the Internet to meet others for sex without condoms are very diverse in their backgrounds, interests and desires.

The strategies the men used were similar to the ones described by marketing researchers on personal ads of newspapers (Hirschman 1987; Jagger 1998). The difference, however, was that in our analysis, the medium of exchange was performances of sexual prowess and of a certain masculine ideal that rejects condoms and that eroticizes risk. Furthermore, it was also clear that, as Dowsett and colleagues(2008) suggested in the context of the same population, sexual objectification becomes a project of the self. It thus also becomes part of these men's online image which, at the same time, provides insight into their desires and into their perception of other men's desires.

Limitations and Conclusions

By sample design, we recruited from highly sexualized websites that did not require membership fees and therefore our results only pertain to men who have sex with men in New York City who use these websites. Consequently, our findings may not be generalizable to all men who use the Internet for bareback sex in the USA or globally. Response bias and having multiple profiles with different characteristics may have interfered with the reporting. However, despite these limitations, this study demonstrates that many men who have sex with men are using Internet profiles to communicate and interact with other men from a wide variety of background, cultures, ethnicities, and countries for bareback sex and therefore represent a population in need of further investigation.

Mixed methods allowed for greater insight into the motivations behind the construction and use of certain online profiles. Because the study was not designed for a direct correspondence between participants from both phases of the study, men whose profiles were sampled on P1 were likely not the same as those recruited and interviewed for P2. However, they all belong to the same pool of men who have sex with men who use specific websites to find partners for bareback sex. Therefore, we believe that this study allowed us to sample an important part of the textual and symbolic performances and interactions that take place online amongst these men. Theexact correspondence between those sampled and those interviewed is not of utmost importance for our analysis.

The data used were collected between 2004 (P1) and 2006 (P2), which means that the realities that we tried to capture may be changing, as new technologies, websites and trends appear and gain popularity, while others fade away. However, the use of profiles in sex websites, including barebacking sites, remains very popular. Therefore, we believe that our insights on the elements present in a profile give a good building stone for future analysis of this reality.

Finally, worthy of notice is the fact that we found a high proportion of HIV-positive men in our sample (P1). This may mean that websites specialising in bareback sex are attracting HIV-positive men to seek partners of the same serostatus, a practice that has been designated“serosorting” (Clatts, Goldsamt, and Yi 2005; Parsons et al. 2006). However, the next-most numerous groupwas that of men who failed to disclose serostatus. These two groups appear to have different patterns of presentation, as HIV-positive men were more likely to include multiple pictures, face pictures, and to mention interest in drug use in their profiles, while non-disclosing men showed the opposite tendency of presentation. As we identified in the interviews with users of these websites, excluding pictures from the profiles may be a strategy to obtain other men's pictures or, alternatively, may suggest a specific form of performance and interaction philosophy, one for which the Internet and the sites used for that purpose should be venues for anonymous sexual encounters. While the frequent presence of pictures of HIV-positive men, as previously discussed, may indicate a sort of inverted perception of stigma in bareback websites, the lack of images in the presentation of non-disclosed HIV status men may indicate that this is a subset of users with higher levels of increased secrecy or anonymity. The implications of the existence of such a sub-group within the online barebacking networks seem well worth investigating.