MEN WHO HAVE SEX WITH men represent a minority of the US male population (estimated at 2.31–132) and most (95%) report having engaged in anal intercourse (AI). Less is known about this practice among heterosexuals,3 although prevalence estimates range between 10% and 35% for heterosexual women in the United States and UK1,2,4,5 and as high as 40% of US men report AI with a women partner.1 In absolute numbers, 7 times more US heterosexual women than homosexual men engage in AI.3 In heterosexual samples, reported AI ranged from 23% of “nonvirgin” university students6 to 32% of sexually active women at high risk of HIV in the previous 6 months7 and is relatively frequent with 7% of sexually active respondents reporting AI at least once a month during the previous year in a household survey.8
Heterosexual AI has been associated with increased HIV transmission within serodiscordant couples in the United States and Brazil9 and with prevalent HIV infection among men and women attending sexually transmitted disease (STD) clinics in India.10 This increased risk is likely due to greater efficiency of transmission and low frequency of condom use during AI. The estimated unadjusted probability of transmitting HIV is 0.08 per contact for receptive AI,11 as compared with 0.001 per coital act for vaginal intercourse.12 Associations between AI and sexually transmitted infections (STIs)9 other than HIV are less consistent. AI has been associated with gonorrhea in the general population but not in STD clinic patients.13 Among women, AI has been significantly associated with abnormal anal cytology,14,15 yet little data exist on rectal carriage of other STI pathogens, largely because women’s ano-rectal sites are rarely tested.
There are few data on factors that may influence the practice of AI among heterosexuals16 or whether AI places them at higher risk for STIs. Such data could help guide interventions to reduce heterosexual HIV transmission via AI, and perhaps clarify whether rectal microbicides for STI prevention among heterosexuals are needed. Therefore, we assessed individual and partnership characteristics associated with AI among young heterosexual individuals attending 3 geographically distinct US STD clinics.
Between November 2001 and May 2004, 1220 adults ages 18 to 26 recruited from waiting rooms in public STD clinics in Seattle, WA, (n = 605), New Orleans, LA (n = 367), and St Louis, MO (n = 248) were interviewed and tested for STIs in the young adults, partnerships, and STD study. Eligibility criteria included presenting with a “new problem” consisting of symptoms of STI, or sexual contact with someone with STI. IRBs at the University of Washington, WA University at St. Louis, and Louisiana State University approved all study procedures, and approval for data analysis was received from the University of California, Los Angeles IRB.
All study subjects participated in a combination of interviewer-and computer-assisted self interviews lasting approximately 1 hour, which included questions from Wave III of the National Longitudinal Study of Adolescent Heath (Add Health)17 as well as more detailed sexual behavior questions. The computer-assisted self interview posed questions about the participant’s 3 most recent romantic and sexual relationships, characteristics of these, sexual experiences, history of and exposure to STI, involvement with the criminal justice system, and substance use. All participants also underwent routine clinical examination with STI screening. Analyses focus on ever having practiced AI and AI with each of the past 3 partners. Because few participants provided complete information on more than 1 partner, our detailed analyses focused on AI reported with the most recent partnership.
Clinical and Laboratory Methods
Clinicians posed questions on sexual history and symptoms, and performed an external genital (men and women) and speculum (women) exam. Men provided a urine specimen and/or urethral swab, while women provided a urine specimen and/or cervical swab, and a self-obtained vaginal swab for STI testing. Mycoplasma genitalium was detected by PCR and transcription mediated assay, and Chlamydia trachomatis and Neisseria gonorrhoeae were detected using culture or nucleic acid amplification tests, as previously described.18–20 Suspected genital herpes (HSV-2) infection was confirmed by viral culture and Trichomonas vaginalis was diagnosed by wet-mount microscopy. “Any STI” was defined as a positive test for any of these pathogens.
In univariate analyses, categorical variables were assessed using Fisher Exact or Pearson chi squared test, whereas continuous variables were assessed using the Wilcoxon rank sum test. Multivariate logistic regression was used to identify factors independently associated with the practice of AI, defined in 1 model as ever having engaged in the practice and in another as having practiced AI with the most recent sex partner. Variables that were associated with STI or AI at P ≤0.05 in univariate analyses were selected for testing in multivariate models and retained if P ≤0.05. All univariate and multivariate analyses were stratified by city using the svy function in Stata version 8.0 (StataCorp, College Station, TX), and separate models were developed for men and women.
Demographic, Sexual Behavior, and Partnership Characteristics
Of the 1220 young adults enrolled, we analyzed data from 1084 individuals who reported no same-sex partners among their last 3 partners, excluding 136 who reported a same sex partner. Among these 1084 heterosexuals, 31.5% were from the New Orleans clinic, 21.3% from the St. Louis clinic, and 47.2% from the Seattle clinic. Most were men (57.7%), 25.7% were nonhispanic white, 62.4% nonhispanic black, 6.2% Hispanic, and 5.8% other. The overall median lifetime number of partners was 9, men reported more partners than women (10 vs. 7, P = 0.13). Similarly, the median number of partners during the past year was higher for men than for women (3 vs. 2, P <0.01). Participants reported a median of 4 sex acts during the past month. At least 1 STI (“Any STI”) was detected in 342/1084 (32%) persons.
Of the 920 participants who provided detailed information on their most recent partner, 201 (22%) reported an age difference of 5 years or more; more common among women (33.6%) than men (12.4%), (P <0.001). Fifty-one percent reported a current partner and among these, an exclusive partner was reported by 271 (57.8%), 91 (19.4%) reported a nonexclusive dating relationship, 48 (10.2%) characterized the partnership as dating infrequently, and 59 (12.6%) said the relationship was only for sex. Among women, 59% reported initiating sex equally, 12% said she initiated sex more; and 29% said the male partner initiated sex. A similar proportion of men (57%) reported initiating sex equally; 31% said that he initiated sex; and 12% said their women partner initiated sex. Overall, 123/587 (21.0%) reported having experienced physical abuse from their most recent partner (i.e., said ‘yes’ to the question “your partner slapped, hit, or kicked you,”) and 70/587 (11.9%) reported being forced to have sex (i.e., said ‘yes’ to the question “your partner insisted on or made you have sex when you didn’t want to.”) More men reported physical abuse (27.3%, 81/297) than women (14.5%, 42/248, P <0.001). Similar percent of men and women reported being forced to have sex (13.1% vs. 10.7%, respectively). Physical abuse and forced sex in the last partnership were significantly associated with reported AI with last partner in men but not women.
Prevalence of AI
Overall 400 (37%) reported ever practicing AI; this was significantly different by gender and STD clinic (Fig. 1) although not for men across different clinics (Table 1). Of those who reported AI, 36% (n = 144) reported condom use during last AI, but this was reported by fewer women (26%) than men (45%) (P <0.001). AI with at least one of the last 3 partners was reported by 266 (28.9%); AI with the last partner was reported by 15.7% of men and 22.6% of women (P = 0.008). Overall, the proportion ever engaging in AI increased steadily with age, going from 23% among 18 year old to 46% of 25 year old (Ptrend <0.001).
Factors Associated With Ever Practicing AI
In univariate analyses, fewer who reported AI reported condom use at last vaginal sex than those who did not practice AI (45.4% vs. 51.9%, P = 0.07). Having more lifetime partners, sex the same day as meeting a partner, and ever exchanging sex for money was significantly more common among those who had ever practiced AI (Table 1). In separate unadjusted logistic regression models for men and women stratified by city, having had sex on the same day as meeting a partner was associated with ever engaging in AI both for men [Odds Ratio (OR) 2.3 (95% CI, 1.55–3.31)] and women [OR 6.0 (3.95–9.10)]. Ever having paid for sex was associated with ever engaging in AI for men [(OR 1.7 (1.04–2.70)], while ever having received money for sex [OR 5.7 (3.13–10.4)] or having greater than 3 lifetime partners [OR 4.9 (2.78–8.58)] were associated with ever engaging in AI for women. Individuals who had ever engaged in AI were more likely to report a history of any STI (65% vs. 57%, P = 0.007), but not more likely to have laboratory-diagnosed STI at study enrollment. History of injection drug use was not associated with ever AI for men, with 4.8% of those reporting AI also reporting drug use versus 2.8% reporting drug use among those not reporting AI, but was significant among women (4.4% vs. 0.4%, respectively), although these numbers are very small.
Separate multivariate analyses stratified by city were conducted for ever having practiced AI for men and women. Women who had sex on the same day they met a partner (AOR 3.9; 95% CI, 2.46–6.21), ever received money for sex (AOR 2.8; 95% CI, 1.40–5.45), and who reported greater than 3 lifetime sex partners (AOR 2.8; 95% CI, 1.56–5.07) were significantly more likely to report ever engaging in AI (Fig. 2). Men who reported having sex on the same day they met a partner (AOR 2.0; 95% CI, 1.33–3.06) were significantly more likely to report ever having AI.
Factors Associated With Engaging in AI With the Most Recent Partner
Of the 920 individuals who provided detailed information on their most recent partner, 173 (19%) reported having had AI at least once with that partner. Among these, 38% (n = 65) reported they had ever used a condom for AI (31% of women and 45% of men, P = 0.06). In univariate logistic regression analysis stratified by city, men who reported having a partnership lasting longer than 3 months [OR 2.22 (1.22–4.04)], using sex toys [OR 7.10 (3.46–14.6)], and having concurrent partners [OR 2.83 (1.64–4.87)] were more likely to report AI with their most recent partner. Among women, AI with the most recent partner was associated with a partnership of longer than 3 months duration [OR 2.24 (1.18–4.25)], having vaginal sex within the first 7 days of the partnership [OR 2.17 (1.28–3.69)], believing that the partner had other concurrent partners [OR 2.33 (1.44–3.78)], having a concurrent partner [OR 2.16 (1.32–3.52)], and using sex toys [OR 5.25 (2.43–11.3)]. Whether the last partnership was exclusive or casual was not associated with engaging in AI within that partnership, nor was laboratory diagnosed STI.
In multivariate logistic regression analysis stratified by site, men who reported sex toy use [AOR 5.6 (2.63–12.0)] and having concurrent partners [AOR 2.2 (1.21–4.11)] in the partnership were more likely to report AI with the most recent partner. For women, having sex within the first 7 days of meeting that partner [AOR 2.4 (1.28–4.37)], believing that their partner had concurrent partners [AOR 1.9 (1.12–3.22)], and using sex toys [AOR 5.7 (2.31–14.0)] were associated with AI in that partnership (Fig. 3).
To consider whether these young adults were having AI instead of vaginal intercourse to preserve their “virginity” we examined AI among those who reported they had never had vaginal intercourse. Of the 20 women who reported never having had vaginal sex, 5 (25%) reported ever engaging in AI, and 4 (20%) reported AI with at least one of the past 3 partners. Of the 46 heterosexual men who reported never having vaginal intercourse, 6 (13%) reported ever having AI, and 7 (15%) reported AI with at least one of the last 3 partners. Male “virgins” were significantly less likely to have had AI than men who had ever had vaginal intercourse (13% vs. 36%, P = 0.001). Similarly, women who were virgins were somewhat less likely to have had AI than women who reported having had vaginal intercourse (25% vs. 42%, P = 0.13).
AI was common among these heterosexual young adults attending STD clinics in all 3 of the cities studied. Condom use during AI was reported by few and by fewer than those reporting condom use during vaginal intercourse. Moreover, the young adults who practiced AI had riskier sexual behavior in general than other young adults. AI was associated with selling/trading sex, having sex with new acquaintances, having concurrent partners, and engaging in other less traditional sexual behaviors such as using sex toys. These associations suggest that those who practice AI have an expanded sexual repertoire.
Except for an association for women in univariate analysis with very small numbers we did not see an association with drug use in our study, yet other studies suggest that drug users are more likely to practice AI than non drug users, particularly methamphetamine users.21,22 We did however; see an association of ever engaging in AI with exchange of sex for money. Although it is not clear if AI is practiced itself for money, it is clearly associated with engagement in commercial sex during which women may engage in sexual practices in response to requests by others. Additionally, both men and women who practiced AI were more likely to have sex with partners they did not know well (i.e., within the same day or week of meeting). Other studies have noted similar associations of AI with trading sex for money and sex with an IDU among young women in non-STD clinic populations.23 Finally, although other studies have noted differences by ethnicity,24 neither the National Survey of Family Growth25 nor our study noted differences in this practice by ethnicity.
Almost one-third of participants in our study reported ever engaging in AI and this was a young group, suggesting that lifetime rates will be high among these STD clinic attendees as they age. Moreover, the high prevalence of this behavior among these young adults suggests an early initiation of this behavior. Our lifetime prevalence of ever practicing AI closely matches that reported in the National Survey of Family Growth (30% of women and 34% of men) which included a much wider age range of 18 to 44 years of age.1 In contrast, AI was reported by only 11% of men 15 to 19 years old in the 1995 National Survey of Adolescent Men,26 and by 21.7% of sexually active women 18 to 29 years in a population based study in low-income neighborhoods in Northern California.23 Despite the different age ranges and sampling frames for these studies, the higher proportion of our young STD patients reporting AI relative to the general population samples of young adults is consistent with previous observations that STD clinic attendees engage in more high risk behaviors.
We found little evidence that AI was being used to maintain virginity as very few virgins reported AI. However, there were few virgins attending these STD clinics and there may be more young adults in the general population who use AI to maintain virginity. Nevertheless, loss of virginity has been linked primarily with vaginal and AI27; thus, it is unlikely that AI is used as much as oral sex to maintain virginity. This suggest that reasons for engaging in AI are more likely social, partnership, and/or physical, rather than to maintain virginity.
Our analyses of AI with the last partner provide unique insight into how AI occurs in partnerships. Unlike other studies23 AI was not associated with the type of partner (i.e., exclusive or casual). However, AI was more common in partnerships where there was concurrency. Thus, AI may be practiced in partnerships that are unstable, and sexual experimentation may be used to sustain such partnerships. The fact that AI was reported more by women who believed their partners had concurrent partnerships and more by men who themselves had concurrent partnerships, lends support to this possibility. Moreover, the partnerships with AI were also partnerships in which sex was initiated quickly (within 7 days), also suggesting that there may be instability or adventurousness in these partnerships as the partners did not know each other well or have an established social relationship before their sexual relationship.
Other studies have noted that AI was reported more often by those engaging in experimental sexual activities such as bisexuality.28 Although our study population was restricted to heterosexuals, in exploratory analyses among the 127 excluded bisexual men, we found that they practiced AI more with their women partners (20%) and in more of their partnerships (42%) than did men who only had sex with women (13%). However, our numbers of bisexual men were too small for more detailed analyses of this issue, warranting further study.
The relatively high proportion of young adults reporting AI raises concerns about increased risk for acquisition and transmission of STIs including HIV among young adults. The practice itself seems to be increasing; the proportion of a random sample of Seattle residents who reported AI increased from 4.3% in 1995 to 8.3% in 2004.29 Although we found no association of AI with a combined indicator of having any of several STIs, we did not test participants for anorectal STI, thus we likely missed cases. The risk for HIV acquisition or transmission via AI is compounded by infrequent condom use during AI, exposure to many partners, exposure to risky partners (including commercial and concurrent partners), and quick initiation of sex after meeting a partner among those practicing AI. These observations suggest that individuals who practice AI may be bridging sexual networks and further increasing the likelihood they will be exposed to or will transmit HIV and highlights the need for broad behavioral counseling for young adults, particularly those seen in STD clinics, about the risks of HIV transmission/acquisition during AI. In our study, condom use with AI was reported by less than half of participants, and fewer women. HIV prevention programs should address the practice of AI with young heterosexuals and encourage condom use to reduce their risk of HIV infection, especially for women where AI has been associated with HIV infection.30
1. Mosher WD, Chandra A, Jones J. Sexual behavior and selected health measures: men and women 15–44 years of age, United States, 2002. Adv Data 2005:1–55.
2. El-Attar SM, Evans DV. Anal warts, sexually transmitted diseases, and anorectal conditions associated with human immunodeficiency virus. Prim Care 1999; 26:81–100.
3. Halperin DT. Heterosexual anal intercourse: prevalence, cultural factors, and HIV infection and other health risks, Part I. AIDS Patient Care STDS 1999; 13:717–730.
4. Evans BA, Bond RA, Macrae KD. Sexual behaviour in women attending a genitourinary medicine clinic. Genitourin Med 1988; 64:43–48.
5. Strang J, Powis B, Griffiths P, et al. Heterosexual vaginal and anal intercourse amongst London heroin and cocaine users. Int J STD AIDS 1994; 5:133–136.
6. Baldwin JI, Baldwin JD. Heterosexual anal intercourse: an understudied, high-risk sexual behavior. Arch Sex Behav 2000; 29:357–373.
7. Gross M, Holte SE, Marmor M, et al. Anal sex among HIV-seronegative women at high risk of HIV exposure. The HIVNET Vaccine Preparedness Study 2 Protocol Team. J Acquir Immun Defic Syndr 2000; 24:393–398.
8. Erickson PI, Bastani R, Maxwell AE, et al. Prevalence of anal sex among heterosexuals in California and its relationship to other AIDS risk behaviors. AIDS Educ Prev 1995; 7:477–493.
9. Guimaraes MD, Munoz A, Boschi-Pinto C, et al. HIV infection among female partners of seropositive men in Brazil. Rio de Janeiro Heterosexual Study Group. Am J Epidemiol 1995; 142:538–547.
10. Rodrigues JJ, Mehendale SM, Shepherd ME, et al. Risk factors for HIV infection in people attending clinics for sexually transmitted diseases in India. BMJ 1995; 311:283–286.
11. Vittinghoff E, Douglas J, Judson F, et al. Per-contact risk of human immunodeficiency virus transmission between male sexual partners. Am J Epidemiol 1999; 150:306–311.
12. Gray RH, Wawer MJ, Brookmeyer R, et al. Probability of HIV-1 transmission per coital act in monogamous, heterosexual, HIV-1-discordant couples in Rakai, Uganda. Lancet 2001; 357:1149–1153.
13. Manhart LE, Aral SO, Holmes KK, et al. Influence of study population on the identification of risk factors for sexually transmitted diseases using a case-control design: the example of gonorrhea. Am J Epidemiol 2004; 160:393–402.
14. Moscicki AB, Hills NK, Shiboski S, et al. Risk factors for abnormal anal cytology in young heterosexual women. Cancer Epidemiol Biomarkers Prev 1999; 8:173–178.
15. Holly EA, Ralston ML, Darragh TM, et al. Prevalence and risk factors for anal squamous intraepithelial lesions in women. J Natl Cancer Inst 2001; 93:843–849.
16. Potterat JJ, Brewer DD, Brody S. Blind spots in the epidemiology of HIV in black Americans. Int J STD AIDS 2008; 19:1–3.
17. Udry JR. The National Longitudinal Study of Adolescent Health (Add Health), Waves I and II, 1994–1996; Wave III, 2001–2002 [machine-readable data file and documentation]. Chapel Hill, NC: Carolina Population Center, University of North Carolina at Chapel Hill; 2003.
18. Dutro SM, Hebb JK, Garin CA, et al. Development and performance of a microwell-plate-based polymerase chain reaction assay for Mycoplasma genitalium
. Sex Transm Dis 2003; 30:756–763.
19. Wroblewski JK, Manhart LE, Dickey KA, et al. Comparison of transcription-mediated amplification and PCR assay results for various genital specimen types for detection of Mycoplasma genitalium
. J Clin Microbiol 2006; 44:3306–3312.
20. Nelson SJ, Manhart LE, Gorbach PM, et al. Measuring sex partner concurrency: it’s what’s missing that counts. Sex Transm Dis 2007; 34:801–807.
21. Zule WA, Costenbader EC, Meyer WJ Jr, et al. Methamphetamine use and risky sexual behaviors during heterosexual encounters. Sex Transm Dis 2007; 34:689–694.
22. Centers for Disease Control and Prevention. Methamphetamine use and HIV risk behaviors among heterosexual men - Preliminary results from five northern California counties, December 2001-November 2003. MMWR Morb Mortal Wkly Rep 2006; 55:273–277.
23. Misegades L, Page-Shafer K, Halperin D, et al. Anal intercourse among young low-income women in California: an overlooked risk factor for HIV? AIDS 2001; 15:534–535.
24. Norris AE, Ford K, Shyr Y, et al. Heterosexual experiences and partnerships of urban, low-income African-Am and Hispanic youth. J Acquir Immun Defic Syndr Hum Retrovirol 1996; 11:288–300.
25. Leichliter J, Chandra A, Liddon N, et al. Prevalence and correlates of heterosexual anal and oral sex in adolescents and adults in the United States. J Inf Dis 2007; 196:1852–1829.
26. Gates GJ, Sonenstein FL. Heterosexual genital sexual activity among adolescent males: 1988 and 1995. Fam Plann Perspect 2000; 32:295–297, 304.
27. Bersamin MM, Fisher DA, Walker S, et al. Defining virginity and abstinence: adolescents’ interpretations of sexual behaviors. J Adolesc Health 2007; 41:182–188.
28. Foxman B, Aral SO, Holmes KK. Heterosexual repertoire is associated with same-sex experience. Sex Transm Dis 1998; 25:232–236.
29. Aral SO, Patel DA, Holmes KK, Foxman B. Temporal trends in sexual behaviors and sexually transmitted disease history among 18- to 39-year-old Seattle, Washington, residents: results of random digit-dial surveys. Sex Transm Dis 2005; 32:710–717.
30. Chirgwin KD, Feldman J, Dehovitz JA, et al. Incidence and risk factors for heterosexually acquired HIV in an inner-city cohort of women: temporal association with pregnancy. J Acquir Immun Defic Syndr Hum Retrovirol 1999; 20:295–299.