Previous studies have demonstrated associations between amphetamine, methamphetamine, or nitrite use and HIV-1 seroconversion,1-5 incident sexually transmitted infections (STIs)6-8 or unprotected anal intercourse (UAI)9-16 among men who have sex with men (MSM). However, the meaning and significance of these associations are unclear. It has been suggested that substance use and sexual risk are markers for risky personality types,17 lack of impulse control,18-20 or desire to escape concern about HIV risk.21-25 In contrast, intoxication with alcohol or illicit substances has been shown to compromise one's ability to negotiate or use condoms,26-30 suggesting a direct association between substance use and UAI. Additionally, after completion of methamphetamine treatment, MSM have reported significant decreases in their number of sexual partners and an increase in their ability to use condoms31,32 compared with just before initiating treatment, suggesting that methamphetamine had impaired their ability to practice safer sex. Higher rates of substance use33,34 and higher HIV incidence35 and prevalence36 among MSM compared with the general population highlight the need to understand associations between substance use and UAI better.
Leigh and Stall37 have proposed that event and within-subjects analyses provide better understanding of the true associations between drug use and HIV/STI risk than studies of global or situational associations. Event and within-subject analyses examine a single instance of sexual behavior and measure the occurrence of substance use at that instance, which provides evidence of a direct association. Within-subjects analyses differ from event analyses by controlling for individual characteristics that are difficult to measure, through using individuals as their own control while comparing instances when risky sexual activity did and did not occur. In contrast, situational association studies measure the frequency of substance use before sexual activity in relation to the occurrence or frequency of high-risk sexual activity. Global association studies examine substance use and high-risk sexual activity in general. These studies are unable to establish temporality (ie, that substance use occurred during or immediately before the high-risk sexual situation) and lack the ability to distinguish between cause and confounding.
Few studies that examine event or within-subjects analyses in the context of substance use and HIV have been published, and some provide conflicting evidence. Among HIV-negative MSM, event level associations between UAI and the use of alcohol, methamphetamine, ecstasy, gamma hydroxyl-butyrate (GHB), or ketamine during sex (but not when used apart from sexual activity)10 have been demonstrated. Additionally, substance use collapsed across amyl nitrites, cocaine, and amphetamine has been associated with serodiscordant UAI in within-subjects analyses among HIV-negative MSM.38 In contrast, an earlier study of MSM with unknown HIV status revealed no within-subjects associations between substance use and UAI.39 Although a growing number of cross-sectional studies have demonstrated global associations between Viagra (sildenafil) use and UAI,9,40-44 to our knowledge, no within-subjects or event studies have taken into account erectile dysfunction medications (EDMs; ie, Viagra®/sildenafil citrate, Levitra®/vardenafil HCL, Cialis®/tadalafil), which represent an emerging class of abused substances.41,44,45
To elucidate the relation between substance use and UAI further, we studied a unique cohort of MSM with recent HIV infection. Occurrence of UAI within this cohort is of particular interest, because recent risk behaviors coincide with acquisition of HIV and further HIV transmission is likely at each episode of UAI after infection because of the high viral loads observed during early infection.46-49 The objectives of this study were to estimate associations between the use of specific recreational substances and EDMs on UAI while using individuals as their own control in within-subjects analyses and to determine if recreational substance use was associated with UAI when considering all participants, including those with no variation in UAI between partners.
Between May 2002 and July 2005, 211 people enrolled in the Acute Infection and Early Disease Research Program (AIEDRP) in San Diego and Los Angeles and provided informed consent. Of those, 207 (98%) completed a computer-assisted self-interview (CASI). All participants had recent HIV infection as determined by one of the following: (1) HIV seroconversion within the previous 12 months (negative HIV enzyme immunoassay [EIA], followed by positive EIA); (2) presence of HIV RNA in plasma but a negative EIA, or (3) results on a detuned EIA that are consistent with early infection. The estimated date of infection for all participants was based on the last HIV-negative test result and serology as previously described.50
This study includes data from 194 of the 207 subjects who were men that reported sexual contact with other men in the previous 12 months. The remaining 13 subjects (4 women, 8 men who reported sexual contact with only women, and 1 man who reported no sexual activity in the previous 12 months) were excluded, because the focus of this study was sexual risk behaviors among MSM. Interview data were collected using Ci3 (Sawtooth Software, Northbrook, IL). The protocol for this study was approved by the Institutional Review Boards of the University of California in San Diego and Los Angeles, Harbor-University of California, Los Angeles Hospital, and Cedar Sinai Hospital in Los Angeles.
Participants were asked to provide detailed information about the last 3 people with whom they had had sexual contact. Questions were asked for each partner regarding types of sexual activities that occurred, substances used just before or during sexual activity, partner demographic information, timing of sexual activity with regard to meeting the partner, partner HIV status, and partner type (eg, main, anonymous). The date of HIV diagnosis was established through reviewing medical records and assigned as the first positive HIV test result that was reported to the participant.
For analyses of cohort characteristics, participant responses regarding UAI with the last 3 partners were grouped by those who reported UAI with none of the last 3 partners, UAI with 1 or 2 (ie, some) of the last 3 partners, and UAI with all the last 3 partners. These 3 groups were compared by demographics, sexual history, and type of substance use using χ2 analysis, the Fisher exact test, and the Kruskal-Wallis test. Trends within these groups were tested using a χ2 test for trend and the Cuzick test for trend.
Repeated measures analyses were used to assess UAI and drug use for each of the last 3 partners individually. Univariate and multivariate conditional logistic regression (CLR) models were used to examine associations between substances used and UAI among the 116 MSM reporting UAI with some of their last 3 partners in within-subjects analyses. CLR enables one to examine change from a single time point to the next at the individual level while conditioning on individual characteristics, thus allowing examination of individual behavior in the presence and absence of substance use while controlling for unmeasured individual characteristics,51 such as personality type.
Analyses that examined associations between UAI and substance use before sexual activity across all 194 participants were conducted using univariate and multivariate generalized linear mixed effects models (GLMMs). GLMMs include within- and between-subjects comparisons, while controlling for repeated measures on the same subject.52 Interactions between EDM use and partner type were examined in GLMMs and CLR models and were considered statistically significant at a P value of 0.1. Analyses were performed using STATA version 8.2 SE (STATA Corporation, College Station, TX).
Patients were interviewed a mean of 16 weeks (median of 14 weeks) after their estimated date of HIV infection and a mean of 5 weeks (median of 3 weeks) after receiving an HIV-positive diagnosis. The mean and median plasma viral loads among participants at enrollment were 5.7 log10 copies/mL for the patients with acute HIV (30 days of infection or less) and 4.6 log10 copies/mL for those with early HIV infection (31 days to 6 months). Respondents had a mean age of 35 years (range: 18-65 years) and were mostly white (69.6%), and almost half (47.9%) reported completing college or higher education (Table 1). Twelve subjects (6.2%) reported sexual activity with men and women in the past 12 months; the remaining 93.8% reported sexual contact exclusively with men. The median reported number of male partners was 99 for lifetime (range: 2-1000 partners), 20.5 in the previous 12 months (range: 1-750 partners), and 4 in the previous 3 months (range: 0-100 partners). Use of any substance with any of the last 3 partners was reported by 58% of participants (Table 1). Use of specific substance types with any of the last 3 partners was reported as follows: methamphetamine in 31%, volatile nitrites in 33%, marijuana in 25%, GHB in 13%, EDM in 23%, and polydrug use (a combination of 2 or more substances, except EDM) in 35%. Few MSM reported use of other substances, including methylenedioxymethamphetamine (MDMA) in 7%, cocaine in 5%, ketamine in 3%, and other drugs in 6% (data not shown).
When categorized by how many of the last 3 partners participants had UAI with, 16% (n = 31) reported no UAI; 60% (n = 116) reported UAI with some but not all of the last 3 partners, and 24% (n = 47) reported UAI with all their last 3 partners. There were no significant differences in demographic or sexual history characteristics across these 3 categories of UAI (Table 1). However, there were significant differences (P < 0.05) by a χ2 or Fisher exact test between these 3 groups by proportion of the use of methamphetamine, volatile nitrites, GHB, and more than 1 substance (polydrug use) with any of the last 3 partners (Table 1). Proportions of drug use increased across categories of UAI from none of the last 3 partners, to some, to all, respectively, for methamphetamine (16.1% vs. 27.6% vs. 51.1%; P = 0.002), nitrites (12.9% vs. 32.8% vs. 44.7%; P = 0.011), polydrug use (16.1% vs. 31.9% vs. 53.2%; P = 0.002), and GHB (3.2% vs. 11.2% vs. 23.4%; P = 0.031) by the χ2 test for trend.
Among the 116 MSM who reported UAI with some (but not all) partners, using CLR, UAI was more commonly reported with partners with whom participants had used methamphetamine (odds ratio [OR] = 5.28), volatile nitrites (OR = 2.55), marijuana (OR = 5.74), multiple classes of substances (polydrug use; OR = 4.18), or any substance (OR = 3.83) when compared with their partners with whom substance use did not occur (Table 2).
Similar associations were seen in multivariate CLR models, which included methamphetamine, nitrites, marijuana, and EDM and controlled for partner type (main vs. all other types), partner age, days between meeting the partner and sexual intercourse, partner HIV status, and whether sexual contact occurred before or after HIV diagnosis (Table 3). In the CLR model, which controls for underlying characteristics of the participants, methamphetamine use was the strongest predictor of UAI (OR = 4.86), followed by marijuana use (OR = 4.01). In addition, a significant interaction between EDM use and partner type was present (P = 0.085), in which EDM use with the main partner (eg, boyfriend, life partner) greatly increased the likelihood of UAI (OR = 13.8). The partner's age and HIV status, partner type, and timing of first sexual contact or sexual contact with regard to HIV diagnosis were not associated with UAI.
Between- and Within-Subjects Analyses
In univariate GLMMs that included the entire sample (N = 194) regardless of variation in UAI between partners, methamphetamine (OR = 3.95), volatile nitrites (OR = 2.67), GHB (OR = 5.42), marijuana (OR = 3.92), polydrug use (OR = 4.30), or use of any substance (OR = 2.71) was associated with UAI (Table 2). In multivariate GLMMs, which included methamphetamine, nitrites, marijuana, and EDM and controlled for partner type, partner age, days between meeting the partner and sexual intercourse, partner HIV status, and whether sexual contact occurred before or after HIV diagnosis, methamphetamine use was the strongest predictor of UAI (OR = 3.52), followed by marijuana use (OR = 2.15) (Table 3). A significant interaction (P = 0.047) between the main partner type and EDM use was also present in (OR = 10.1). In addition, having a main partner (OR = 2.55) or an HIV-positive partner (OR = 2.23) or sexual contact occurring before HIV diagnosis (OR = 1.75) was also associated with UAI. No additional variables were associated with UAI.
In this study of MSM with recent HIV infection, methamphetamine, marijuana, and EDM use with main partners was associated with UAI in within-subjects analyses (using CLR), suggesting that the use of these substances may increase UAI independent of individual characteristics and that this association might not be confounded by personality for these drugs as has been previously suggested.17-19 Additionally, in GLMMs that examined within- and between-subjects differences in the entire sample, use of methamphetamine, marijuana, or EDM with a main partner was associated with UAI, suggesting that the results seen in within-subjects analyses also may exist between subjects; however, data were too sparse to test this formally as previously suggested.53 The current study contributes to the overall understanding of drug use and UAI by (1) providing support to prior studies that demonstrate associations between methamphetamine or EDM and sexual risk behavior,9,10,38,54,55 (2) clarifying that a direct association is likely to exist between specific drugs and UAI, and (3) providing evidence that the use of methamphetamine, EDM, and possibly other illicit substances may contribute to HIV transmission.
In all analyses, the most important predictor of UAI among the last 3 sexual partners was methamphetamine use, suggesting that methamphetamine is an independent predictor of HIV transmission. To our knowledge, our study is the first to demonstrate that methamphetamine is associated with UAI among recently HIV-infected MSM, while controlling for individual factors using within-subjects analyses. Considering the high transmissibility of HIV during early infection, these analyses suggest that methamphetamine may contribute significantly to HIV transmission from newly infected MSM to others. The median number of days for the last sexual contact with the last 3 partners was 7 days for the last partner, 30 for the second to last partner, and 45 for the partner proceeding the second to last. Based on calculations for estimated date of infection, we estimate that 391 of the 572 reported partners were sexual partners after HIV infection occurred. UAI was reported with 55% (n = 214) of these partners, and among these, 82% were believed to uninfected with HIV (n = 84) or of unknown HIV status (n = 92). Additionally, viral loads at study entry were fairly high because of acute infection (mean viral load of 5.7 log10 copies/mL) or early infection (mean viral load of 4.6 log10 copies/mL), which may also increase the risk of HIV transmission. These data raise concern that substance use may contribute to onward transmission of HIV among recently infected MSM. A modest reduction in UAI was reported with partners after HIV diagnosis, from 59.7% to 50.7% of partners, but substance use persisted after diagnosis.
There are many plausible pathways by which methamphetamine could increase the risk of UAI. Methamphetamine could impair judgment or reduce the ability to negotiate condom use through direct effects on mental functioning.56 Additionally, methamphetamine has been reported to increase individuals' desire for sexual activity,57-60 which, independently or in combination with modified mental functioning, could result in increased likelihood of UAI.
Of interest was our finding that the association between UAI and EDM use differed by partner type. EDM use was only associated with UAI when used with main partners but not with other partner types. To our knowledge, no study has documented a differential effect of EDM use on UAI in relation to partner type. However, studies examining nitrites, marijuana, and GHB10 or alcohol26,27 use have demonstrated differences in sexual behavior by partner type. It is possible that the association we observed reflects the differential use of EDM according to the specific type of anal intercourse (insertive or receptive). EDMs are more likely to be used by the insertive rather than the receptive partner during UAI, and most participants in our sample reported being the latter. Among the 102 participants for whom the type of UAI (insertive vs. receptive) was measured, only 30% reported being the insertive partner with their last 3 partners. Additionally, participants were more likely to report insertive UAI with main partners than other types (OR = 1.98; P = 0.02) by GLMMs, suggesting that the interaction between EDM use and main partners could be attributable to sexual positioning. Data were too sparse to determine if EDM use was independently associated with insertive UAI, however. Regardless of sexual positioning, the relatively high use of EDMs among MSM who do not have erectile dysfunction, as demonstrated here and in previous studies,41,44 does raise concern about EDM misuse and the contribution of misuse to HIV transmission.
The association between nitrite use and UAI was only marginally significant in GLMM and CLR analyses, which contradicts the findings of previous studies that demonstrated associations between nitrite use and UAI11-15 or HIV seroconversion.2,4,5,61 In our study, the association was not as strong as for methamphetamine, even though the prevalence of use was similar, suggesting that nitrites may be less likely to contribute directly to UAI than other substances. Lack of association between UAI and nitrite use has been demonstrated previously,10 and frequent use of nitrites (once per week or more) was not associated with increased risk of UAI with a serodiscordant partner.38 The ambiguity of the association between nitrite use and UAI may be a result of the use of nitrites for preplanned UAI, because nitrites do not seem to alter mental functioning.62
Significant associations were also observed between marijuana use and UAI among those who reported variation in UAI and the overall sample, suggesting that marijuana is an independent risk factor for UAI, but no trend of increasing use with UAI category was observed. In contrast to our study, a previous event analysis study demonstrated that marijuana use was associated with UAI in general but not when marijuana was used during sexual activity,10 suggesting that marijuana use is more likely to be a marker of high-risk behaviors in general than a risk factor for UAI. Although the associations between marijuana use and UAI in our study may be valid, more than half of marijuana users also used methamphetamine, suggesting that associations between UAI and marijuana may be an artifact of the overlap in the use of methamphetamine. Overlap between marijuana or nitrite use and methamphetamine use is not uncommon among MSM;63 therefore, future studies could benefit from examining marijuana and nitrite use in within-subjects analyses and take into account polydrug use.
As with all studies, our study has some inherent limitations. Many studies have suggested that alcohol use is associated with increased UAI among HIV-infected64-66 and HIV-uninfected MSM.10,13,67,68 Alcohol use was not measured in our study; therefore, associations between UAI and alcohol or interactions between alcohol and substance use could not be examined. Our sample may not be representative of all MSM who are at risk for HIV because they were only sampled from southern California, most self-reported white ethnicity, and they were well educated. Additionally, substance use may vary by region, and MSM on the east coast of the United States may be more likely to use substances such as heroin or cocaine, whereas those on the west coast may be more likely to use substances such as methamphetamine.33,69 Therefore these data should not be extrapolated to all types of substance use. Additionally, it would be helpful to examine differences in within- and between-subjects effects of substance use on UAI;53 however, to do so, a larger sample would be required.
These analyses provide new evidence supporting a growing body of literature demonstrating that use of methamphetamine just before or during sexual activity increases the likelihood of UAI among MSM and raise concern about substance use increasing the risk of HIV transmission. Additionally, our study demonstrates that this is occurring among recently infected MSM, suggesting that methamphetamine use may be helping to propagate the current HIV epidemic among MSM in the United States. Interventions that focus on methamphetamine abuse prevention or rehabilitation and help to prevent the adoption of emerging substances of abuse among recently HIV-infected MSM may help to reduce HIV incidence among MSM.
The authors acknowledge and thank W. Susan Cheng for her assistance in data collection and management and Tari Gilbert, Jacqui Pitt, Paula Potter, Joanne Santangelo, and the University of California, San Diego Antiviral Research Center staff for their support in data collection. They also thank Dr. Simon Frost for his suggestions and assistance in developing and writing this manuscript and 2 anonymous reviewers for their helpful suggestions in improving the quality of this manuscript. Most of all, they thank the participants for the time and effort they have contributed to this study; without their willingness to participate and share their personal information with us, this research would not be possible.
1. Burcham JL, Tindall B, Marmor M, et al. Incidence and risk factors for human immunodeficiency virus seroconversion in a cohort of Sydney homosexual men. Med J Aust
2. Page-Shafer K, Veugelers PJ, Moss AR, et al. Sexual risk behavior and risk factors for HIV-1 seroconversion in homosexual men participating in the Tricontinental Seroconverter Study, 1982-1994. Am J Epidemiol
3. Chesney MA, Barrett DC, Stall R. Histories of substance use and risk behavior: precursors to HIV seroconversion in homosexual men. Am J Public Health
4. Buchbinder S, Vittinghoff E, Heagerty P, et al. Sexual risk, nitrite inhalant use, and lack of circumcision associated with HIV seroconversion in men who have sex with men in the United States. J Acquir Immune Defic Syndr
5. Ostrow DG, DiFranceisco WJ, Chmiel JS, et al. A case-control study of human immunodeficiency virus type 1 seroconversion and risk-related behaviors in the Chicago MACS/CCS Cohort, 1984-1992. Multicenter AIDS Cohort Study. Coping and Change Study. Am J Epidemiol
6. Hirshfield S, Remien RH, Walavalkar I, et al. Crystal methamphetamine use predicts incident STD infection among men who have sex with men recruited online: a nested case-control study. J Med Internet Res
7. Cowan RL, Lyoo IK, Sung SM, et al. Reduced cortical gray matter density in human MDMA (Ecstasy) users: a voxel-based morphometry study. Drug Alcohol Depend
8. Tabet SR, Krone MR, Paradise MA, et al. Incidence of HIV and sexually transmitted diseases (STD) in a cohort of HIV-negative men who have sex with men (MSM). AIDS
9. Colfax GN, Mansergh G, Guzman R, et al. Drug use and sexual risk behavior among gay and bisexual men who attend circuit parties: a venue-based comparison. J Acquir Immune Defic Syndr
10. Rusch M, Lampinen TM, Schilder A, et al. Unprotected anal intercourse associated with recreational drug use among young men who have sex with men depends on partner type and intercourse role. Sex Transm Dis
11. Mulry G, Kalichman SC, Kelly JA. Substance use and unsafe sex among gay men-global versus situational use of substances. J Sex Educ Ther
12. Strathdee SA, Hogg RS, Martindale SL, et al. Determinants of sexual risk-taking among young HIV-negative gay and bisexual men. J Acquir Immune Defic Syndr Hum Retrovirol
13. Woody GE, Donnell D, Seage GR, et al. Non-injection substance use correlates with risky sex among men having sex with men: data from HIVNET. Drug Alcohol Depend
14. Clutterbuck DJ, Gorman D, McMillan A, et al. Substance use and unsafe sex amongst homosexual men in Edinburgh. AIDS Care
15. Ekstrand ML, Stall RD, Paul JP, et al. Gay men report high rates of unprotected anal sex with partners of unknown or discordant HIV status. AIDS
16. Morin SF, Steward WT, Charlebois ED, et al. Predicting HIV transmission risk among HIV-infected men who have sex with men. J Acquir Immune Defic Syndr
17. Semple SJ, Patterson TL, Grant I. Psychosocial predictors of unprotected anal intercourse in a sample of HIV positive gay men who volunteer for a sexual risk reduction intervention. AIDS Educ Prev
18. Ross MW, Mattison AM, Franklin DR Jr. Club drugs and sex on drugs are associated with different motivations for gay circuit party attendance in men. Subst Use Misuse
19. Mccoul MD, Haslam N. Predicting high risk sexual behaviour in heterosexual and homosexual men: the roles of impulsivity and sensation seeking. Pers Individ Dif
20. Patterson TL, Semple SJ, Zians JK, et al. Methamphetamine-using HIV-positive men who have sex with men: correlates of polydrug use. J Urban Health
21. Rhodes T, Stimson GV. What is the relationship between drug taking and sexual risk? Social relations and social research. Sociol Health Illn
22. Rhodes T. Culture, drugs and unsafe sex: confusion about causation. Addiction
23. McKirnan DJ, Ostrow DG, Hope B. Sex, drugs and escape: a psychological model of HIV-risk sexual behaviours. AIDS Care
24. Ostrow D, McKirnan D. Prevention of substance-related high-risk sexual behavior among gay men: critical review of the literature and proposed harm reduction approach. J Gay Lesbian Med Assoc
25. McKirnan DJ, Vanable PA, Ostrow DG, et al. Expectancies of sexual “escape” and sexual risk among drug and alcohol-involved gay and bisexual men. J Subst Abuse
26. Leigh BC. Alcohol and condom use-a meta-analysis of event-level studies. Sex Transm Dis
27. Vanable PA, McKirnan DJ, Buchbinder SP, et al. Alcohol use and high-risk sexual behavior among men who have sex with men: the effects of consumption level and partner type. Health Psychol
28. Wingood GM, DiClemente RJ. The influence of psychosocial factors, alcohol, drug use on African-American women's high-risk sexual behavior. Am J Prev Med
29. Tawk HM, Simpson JM, Mindel A. Condom use in multi-partnered males: importance of HIV and hepatitis B status. AIDS Care
30. Stone E, Heagerty P, Vittinghoff E, et al. Correlates of condom failure in a sexually active cohort of men who have sex with men. J Acquir Immune Defic Syndr
31. Reback CJ, Larkins S, Shoptaw S. Changes in the meaning of sexual risk behaviors among gay and bisexual male methamphetamine abusers before and after drug treatment. AIDS Behav
32. Shoptaw S, Reback CJ, Frosch DL, et al. Stimulant abuse treatment as HIV prevention. J Addict Dis
33. Thiede H, Valleroy LA, MacKellar DA, et al. Regional patterns and correlates of substance use among young men who have sex with men in 7 US urban areas. Am J Public Health
34. Beckett M, Burnam A, Collins RL, et al. Substance use and high-risk sex among people with HIV: a comparison across exposure groups. AIDS Behav
35. Centers for Disease Control and Prevention. Increases in HIV diagnoses-29 states, 1999-2002. MMWR Morb Mortal Wkly Rep
36. Quan VM, Steketee RW, Valleroy L, et al. HIV incidence in the United States, 1978-1999. J Acquir Immune Defic Syndr
37. Leigh BC, Stall R. Substance use and risky sexual behavior for exposure to HIV: issues in methodology, interpretation, and prevention. Am Psychol
38. Colfax G, Vittinghoff E, Husnik MJ, et al. Substance use and sexual risk: a participant- and episode-level analysis among a cohort of men who have sex with men. Am J Epidemiol
39. Gillmore MR, Morrison DM, Leigh BC, et al. Does “high=high risk”?: an event-based analysis of the relationship between substance use and unprotected anal sex among gay and bisexual men. AIDS Behav
40. Chu PL, McFarland W, Gibson S, et al. Viagra use in a community-recruited sample of men who have sex with men, San Francisco. J Acquir Immune Defic Syndr
41. Crosby R, DiClemente RJ. Use of recreational Viagra among men having sex with men. Sex Transm Infect
42. Kim AA, Kent CK, Klausner JD. Increased risk of HIV and sexually transmitted disease transmission among gay or bisexual men who use Viagra, San Francisco 2000-2001. AIDS
43. Sherr L, Bolding G, Maguire M. Viagra use and sexual risk behavior among gay men in London. AIDS
44. Paul JP, Pollack L, Osmond D, et al. Viagra (sildenafil) use in a population-based sample of U.S. men who have sex with men. Sex Transm Dis
45. Aldridge J, Measham F. Sildenafil (Viagra) is used as a recreational drug in England. BMJ
46. Pilcher CD, Tien HC, Eron JJ, et al. Brief but efficient: acute HIV infection and the sexual transmission of HIV. J Infect Dis
47. Little SJ, McLean AR, Spina CA, et al. Viral dynamics of acute HIV-1 infection. J Exp Med
48. Fiebig EW, Wright DJ, Rawal BD, et al. Dynamics of HIV viremia and antibody seroconversion in plasma donors: implications for diagnosis and staging of primary HIV infection. AIDS
49. 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
50. Little SJ, Holte S, Routy JP, et al. Antiretroviral-drug resistance among patients recently infected with HIV. N Engl J Med
51. Breslow NE, Day NE. The analysis of case-control studies. In: Statistical Methods in Cancer Research
, vol. 1. Lyon, France: IARC Scientific Publications; 1980.
52. Neuhaus JM, Kalbfleisch JD, Hauck WW. A comparison of cluster-specific and population-averaged approaches for analyzing correlated binary data. Int Stat Rev
53. Neuhaus JM, Kalbfleisch JD. Between- and within-cluster covariate effects in the analysis of clustered data. Biometrics
54. Colfax G, Coates TJ, Husnik MJ, et al. Longitudinal patterns of methamphetamine, popper (amyl nitrite), and cocaine use and high-risk sexual behavior among a cohort of San Francisco men who have sex with men. J Urban Health
55. Chmiel J, Detels R, Kaslow RA, et al. Factors associated with prevalent HIV infection in the multicenter AIDS cohort. Am J Epidemiol
56. McKim WA. Drugs and Behavior
. 5th ed. Upper Saddle River: Prentice Hall, New Jersey; 2003.
57. Green AI.“Chem friendly”: the institutional basis of “club drug” use in a sample of urban gay men. Deviant Behav
58. Buffum J. Substance abuse and high-risk sexual behavior: drugs and sex-the dark side. J Psychoactive Drugs
59. Semple SJ, Patterson TL, Grant I. Motivations associated with methamphetamine use among HIV+ men who have sex with men. J Subst Abuse Treat
60. Gorman EM, Barr BD, Hansen A, et al. Speed, sex, gay men, and HIV: ecological and community perspectives. Med Anthropol Q
61. Chesney MA, Barrett DC, Stall R. Histories of substance use and risk behavior: precursors to HIV seroconversion in homosexual men. Am J Public Health
62. French RS, Power R. Self-reported effects of alkyl nitrate use: a qualitative study amongst targeted groups. Addict Res
63. Patterson TL, Semple SJ, Zians JK, et al. Methamphetamine-using HIV-positive men who have sex with men: correlates of polydrug use. J Urban Health
64. Parsons JT, Kutnick AH, Halkitis PN, et al. Sexual risk behaviors and substance use among alcohol abusing HIV-positive men who have sex with men. J Psychoactive Drugs
65. Poppen PJ, Reisen CA, Zea MC, et al. Predictors of unprotected anal intercourse among HIV-positive Latino gay and bisexual men. AIDS Behav
66. Purcell DW, Parsons JT, Halkitis PN, et al. Substance use and sexual transmission risk behavior of HIV-positive men who have sex with men. J Subst Abuse
67. Hirshfield S, Remien RH, Humberstone M, et al. Substance use and high-risk sex among men who have sex with men: a national online study in the USA. AIDS Care
68. Ruiz J, Facer M, Sun RK. Risk factors for human immunodeficiency virus infection and unprotected anal intercourse among young men who have sex with men. Sex Transm Dis
69. Sullivan PS, Nakashima AK, Purcell DW, et al. Geographic differences in noninjection and injection substance use among HIV-seropositive men who have sex with men: western United States versus other regions. J Acquir Immune Defic Syndr Hum Retrovirol