Viagra use and sexual risk behaviour among gay men in London

The use of viagra (sildenafil), with or without a prescription, may have implications for safer sexual practice and HIV prevention 1. We investigated this among gay men using gyms in central London as part of an ongoing investigation of sexual risk behaviour. All gay/bisexual men using one of five gyms in central London in March–April 1999 were asked to complete an anonymous questionnaire concerning HIV status, unprotected anal intercourse (UAI) in the previous 3 months, HIV status of UAI partner(s), use of anabolic steroids and other recreational drugs in the previous 6 months. Men were asked if they had ever taken sildenafil (prescription or recreational), whether they had taken it in the previous 3 months and, if so, whether they had had UAI when they took sildenafil. Questionnaires were distributed in each gym over a 6 day period. Of the five gyms, one was exclusively gay whereas the others estimated that gay men comprised 40–90% of their membership. The response rate was calculated using these estimates. Details of the methods have been reported elsewhere 2,3. Completed questionnaires were returned by 720 gay/bisexual men (estimated response rate 40–50%); 40 men had not had sex with another man in the previous 6 months, and for three, data were missing. Of the remaining 677 men, 100 (14.8%) had ever taken sildenafil, 82 more than once and 63 in combination with other recreational drugs. HIV-positive men were significantly more likely to have ever taken sildenafil (27.1%; 29/107) than HIV-negative (14.2%; 61/431) or never-tested men (7.2%; 10/139) (P < 0.001). Only 17 men had been prescribed the drug whereas the remaining 83 had taken it recreationally. The ratio of prescription to recreational use did not vary significantly by HIV status (1 : 5 overall). Of the 29 HIV-positive ever-users, 17 (59%) were on HIV triple combination therapy. For both HIV-positive and HIV-negative men, there were no significant differences between men who had ever taken or had never taken sildenafil in median age, employment status, relationship status, having been paid for sex or having had a sexually transmitted disease in the previous 6 months. However, ever-users were significantly more likely than never-users to report serodiscordant UAI (not necessarily while taking sildenafil) or UAI with a partner of unknown status (HIV-negative men only). Ever-users were also significantly more likely to have used recreational drugs or anabolic steroids in the previous 6 months (Table 1). Controlling for steroid use among HIV-negative men yielded P values of 0.002 for status-unknown UAI and 0.07 for discordant UAI in relation to ever-using sildenafil.Table 1: Frequency of unprotected anal intercourse, anabolic steroid and recreational drug use among gay men. HIV-negative men who had taken sildenafil in the previous 3 months (n = 50) were significantly more likely to report serodiscordant UAI while on sildenafil than HIV-negative never-users (8.0 versus 1.4%, P < 0.01) (Table 1). This differential remained significant after further controlling for steroid use (P = 0.03). No such association was seen for HIV-positive men who had taken sildenafil in the previous 3 months (n = 25). Data were not analysed for never-tested men because only 10 had ever taken sildenafil. Overall, one in seven gay men surveyed in central London gyms had taken sildenafil. The overwhelming majority had taken it recreationally rather than on prescription. There was a twofold difference in reported use between HIV-positive and HIV-negative men (27 versus 14%). The elevated frequency among HIV-positive men may partly reflect potency problems associated with their infection, although in all groups most men had taken the drug recreationally. More than half the HIV-positive men who had taken sildenafil were on triple combination therapy that probably included a protease inhibitor. A potential interaction between protease inhibitors and sildenafil has been reported 4, something which all HIV-positive men should be made aware of. The use of sildenafil was associated with high-risk sexual behaviour (not necessarily while taking the drug) as well as with the use of recreational drugs or anabolic steriods. This suggests that some men may have added sildenafil to their risk-taking repertoire, rather than sildenafil per se leading to an increase in their risk behaviour. Nonetheless, sildenafil may have led to high-risk sexual behaviour among some HIV-negative men, who were more likely to report serodiscordant UAI while taking sildenafil than HIV-negative men who had never used the drug. Being a cross-sectional survey, our investigation could not establish a causal pathway. The study population was not a random sample and the response rate was lower than that reported in other surveys of gay men 5. For these reasons, the emergence of sildenafil as a recreational drug among London gay men reported here merits further investigation in other community samples. Lorraine Sherra Graham Boldinga Mark Maguireb Jonathan Elforda