The prevalence of anal intercourse (AI) is well characterized among men who have sex with men (MSM), however, less is known about this practice among heterosexuals. Recent studies suggest that AI is a commonly practiced behavior with opposite sex partners and the prevalence has increased in recent years. Lifetime occurrence of AI among heterosexuals in the United States has been reported by 30% to 33% of respondents in a national survey and by 23% to 37% of young adults and college students.1–5 The prevalence of recent experience with heterosexual AI is also relatively high and ranges from 18% to 22% of clients attending sexually transmitted disease (STD) clinics and 5% to 42% of substance using women.6–9 Furthermore, the percentage of men and women who report engaging in heterosexual AI, including unprotected AI has increased since the early 1990s.2,6,10
Most studies that have examined a relationship with AI and race/ethnicity have found no association,4,8,11,12 although there is some evidence to suggest that whites are more likely to engage in AI.1,9 Other factors found to be associated with heterosexual AI include substance use, injection drug use, number of sex partners, exchanging drugs or money for sex, and condom use.1,3,4,7–9,11,12 In studies where condom use was assessed, the majority of participants reported never using condoms for anal sex, suggesting increased likelihood of exposure to sexually transmitted infections (STI), including the human immunodeficiency virus (HIV).7,8,11 While a number of studies have shown that AI is an independent predictor of HIV among both men and women, data on associations between AI and STIs other than HIV are less clear.5,13–17 This is largely because women are rarely tested for anorectal STIs, and studies which report on STIs are almost entirely based on self-reported lifetime history, with few studies conducting STI testing concurrent with the timeframe for which anal intercourse is assessed.
Although it is clear that AI is a relatively common sexual practice, which is increasing in prevalence6,10 and has a high probability of transmission of STIs, data on correlates of AI are mixed. Furthermore, information on the prevalence of STIs among heterosexuals reporting recent anal intercourse is limited. Therefore, we sought to identify demographic and behavioral correlates of AI, as well as associations with STI outcomes among a diverse group of clients attending public STD clinics. We hypothesize that the prevalence of recent heterosexual AI will vary by race/ethnicity and risk behaviors and that factors associated with AI will be different by race/ethnicity.
MATERIALS AND METHODS
Study Population and Design
We conducted a cross-sectional study of clients attending 13 public STD cinics in Los Angeles County, CA. Women and heterosexual men were eligible for inclusion. Women who reported having only female sex partners were excluded. Heterosexual men were defined as men reporting only female sex partners in the past 12 months. Men who reported same sex Qäpartners were excluded regardless of whether they also reported female partners, because clients reported on behaviors overall and not in the context of specific sexual partners. Therefore, we were unable to determine if AI was practiced with female or male partners. Data were collected for new visits, which was defined as an initial visit for a new complaint, on all eligible clients attending the public STD clinics between May 2006 and December 2007. The study was approved by the Human Subjects review committee at the University of California Los Angeles.
The data for this analysis were abstracted from existing medical record and data collection systems in place at the STD clinics. Electronic risk assessment and intake data included information on demographics, types of sexual contact, substance use, and other risk factors. The assessments were conducted by the nurse or other trained public health staff at the time of the clinic visit, using face-to-face interviews. Specifically, clients were routinely questioned about the type of sexual contact, including anal receptive and insertive sex in the last 90 days before their current visit. They were also asked to report on the gender of sex partners in the past 12 months and whether any of these partners were anonymous partners, injection drug users, or incarcerated. In addition, clients reported whether in the past 12 months they: (1) exchanged drugs/money for sex, (2) were incarcerated, and/or (3) reported substance use including alcohol, cocaine, ecstasy, gamma hydroxybutyrate, heroin, ketamine, marijuana, methamphetamine, nitrates, or used needles to inject drugs. Substance use, including alcohol use, was assessed as any use in the past 12 months and particularly in the case of alcohol the amount, whether it included binge drinking or was in the context of sexual activity was not assessed. Given the potential for misclassification of social consumption of alcohol as a substance of abuse, we used two separate variables for our analyses including alcohol use and a composite variable of any substance use except for alcohol.
All clients seen for new visits were routinely tested for urogenital chlamydia and gonorrhea, using urine-based nucleic acid amplification testing Aptima Combo 2 (GenProbe, San Diego, CA). Chlamydia and gonorrhea rectal screening of women reporting AI was not routinely performed, with 19% being screened for rectal chlamydia and 46% for rectal gonorrhea. The difference in the prevalence of rectal screening resulted from differences in the availability of an approved test. Chlamydia was defined as any case of urogenital or rectal infection, while gonorrhea was defined as any case of urogenital, pharyngeal, or rectal infection. Furthermore, clients were routinely tested for syphilis and offered HIV testing. Health department disposition for those testing positive for syphilis was obtained in order to determine cases of early syphilis (i.e., primary, secondary, and early latent). In addition, those who did not receive an HIV test during their visit, but were known positives were included in the HIV+ group.
Given the differences in sexual behavior by gender, we conducted separate analyses for heterosexual men and women. We determined the prevalence of heterosexual AI by gender and for all demographic, risk behavior characteristics, and STI results. Differences between groups were evaluated using t-tests and Mantel-Haenszel chi-square methods as appropriate. Because clients could have repeated visits (i.e., more than one new visit during the study period), we used hierarchical regression models, using generalized estimating equations in order to account for within subject correlations.18,19 Variables tested for inclusion in the multivariate models were based on univariate analyses and factors previously associated with AI. To be included in the final model, all covariates considered were required to be significant (α = 0.05) or specified a priori as risk factors based on the existing literature. Deviance tests and score statistics for type 3 generalized estimating equations analysis were used to determine the best model. In addition, separate models were analyzed for each race/ethnic group to determine whether the association between risk behaviors and AI was modified across levels of race/ethnicity.
Characteristics of Study Population
Between May 2006 and December 2007 more than 32,000 unique clients were seen for new visits and a total of 28,184 (87%) were eligible and included in this analysis. We excluded 2844 (9%) men who reported male partners (exclusively or in combination with female partners) and 184 (0.6%) women who reported having only female partners. The eligible clients resulted in 34,791 new clinic visits, with 23,651 clients having only one visit and 4566 clients having two or more visits (range: 2–10) during the study period.
Women comprised 39% (n = 10,933) of the study population. The mean age of female participants was 30.3 years (standard deviation [SD]: 11.2), with nearly 60% of women being 30 years or younger. Most women were either African American (47%) or Hispanic (35%), with 11% identifying as white. Based on all visits by women eligible and included in this analysis (n = 13,695), more than one-quarter (27%) reported some type of substance use in the past 12 months, with marijuana being the most common substance reported (23%), followed by cocaine (4%), and methamphetamine (3%). In terms of sexual risk behaviors in the 12 months before the visit, 4% reported having anonymous sex, 4% reported being incarcerated, and 2% reported exchange of money or drugs for sex. The prevalence of STIs was relatively high, with 17% (n = 2245) being diagnosed with at least one STI at the time of their clinic visit. Among the 1780 cases of chlamydia, the majority were based on urogenital infections (n = 1774) and a small number of rectal infections (n = 34). Similarly, the majority of the 617 cases of gonorrhea infection were based on urogenital infections (n = 508), followed by pharyngeal infections (n = 190), and rectal infections (n = 38).
Male clinic clients included in this analysis (n = 17,251) were slightly older than female clients with a mean age of 32.5 years (SD 11.1). African Americans comprised the single largest ethnic group (41%), followed by Hispanics (36%), and whites (16%). Across all visits (n = 21,056) the prevalence of any substance use was 37%, with 33% reporting marijuana use in the past 12 months, 5% reporting cocaine use, and 3% reporting methamphetamine use. Other risk factors included 10% reporting anonymous sex, 8% reporting incarceration, and 2% reporting exchange of money or drugs for sex. The proportion of men diagnosed with at least one STI at the time of their clinic visit was 17% (n = 3453), with more chlamydia infections (n = 2451) than gonorrhea infections (n = 1272).
Prevalence of Heterosexual Anal Intercourse
Overall, 10% of heterosexual men (n = 1978) and 10% of women (n = 1364) reported AI with an opposite sex partner đin the 90 days preceding their clinic visit. Among these clients the prevalence of the different types of sexual contact was similar for both men and women such that 98% reported vaginal intercourse, 87% reported oral sex, 6% reported rimming, and almost no one reported practicing AI exclusively (<0.01%).
We also found that the prevalence of AI varied by race/ethnicity (Fig. 1). Among women, the prevalence was highest among white and Hispanic women (14% and 13%, respectively) and lowest among African American women (8%, P < 0.01). For men, AI was highest among Hispanic men (14%) and lowest among Asian (6%) and African American men (7%, P < 0.01). Furthermore, the prevalence of AI varied by behavioral factors. Women who reported both male and female partners had a higher prevalence of AI as compared to those who reported having only male partners (16.5% vs. 10%; P < 0.01) (Table 1). For both men and women those who reported exchange of drugs or money for sex, anonymous sex, injection drug use or sex with an injection drug user, were more likely to report AI, however, the prevalence varied by race/ethnicity (Tables 1 and 2). For instance, the prevalence of AI was highest among women who reported exchange of drugs or money for sex with 40.6% of white women, 37.5% of Hispanic women, and 32.0% of African American women in this group reporting AI (Table 1). Among men, those who reported sex with an injection drug user were more likely to report AI as compared to those who did not report sex with an injection drug user (25.5% and 9.6%, respectively; P < 0.01) (Table 2). However, this difference was primarily noted among white men (37.0% vs. 9.7%; P < 0.01) and Hispanic men (35.7% vs. 13.2%; P < 0.01) and no differences were noted among African American men (6.1% vs. 7.1%; P = 0.82). The prevalence of AI was also substantially higher among both men and women who reported substance use. For instance, more than 20% of men and women who reported methamphetamine use reported AI as compared to 10% among nonusers (P < 0.01). The pattern of substance use and its association with AI also varied by race/ethnicity.
Factors Associated With Anal Intercourse Among Women
Based on multivariate analysis African American women were less likely to report AI as compared to white women (adjusted odds ratio [AOR] = 0.53; 95% confidence interval [CI]: 0.43–0.65; Table 3). Overall, we also found that women who reported having both male and female sex partners, exchange of drugs or money for sex, and substance use were more likely to report experience with recent AI (Table 3). Factors associated with AI also varied by race/ethnicity. Among African American women, those who reported exchange of drugs or money for sex (AOR = 2.98; 95% CI: 1.83–4.84) and anonymous sex (AOR = 1.71; 95% CI: 1.13–2.59) were more likely to report AI than those who did not report these behaviors. However, STI diagnosis including chlamydia, gonorrhea, early syphilis, or HIV was not associated with AI. Among Hispanic women, those who reported injection drug use were more likely to report recent AI as compared to those who did not report injection drug use. Furthermore, those who were diagnosed with gonorrhea were nearly twice as likely to report AI as compared to those who were not diagnosed with gonorrhea (AOR = 1.74; 95% CI: 1.04–2.91). STI diagnoses were also associated with AI among white women, with those diagnosed with gonorrhea being nearly four times as likely to report recent AI (AOR = 3.81; 95% CI: 1.66–8.94) and those diagnosed with early syphilis more than five times as likely to report recent AI (AOR = 5.31; 95% CI: 4.15–6.81).
Factors Associated With Anal Intercourse Among Men
Among men, after adjustment for other factors, African American men were less likely to report AI (AOR = 0.70; 95% CI: 0.60–0.82), while Hispanic men were more likely to report recent AI with opposite sex partners (AOR = 1.50; 95% CI: 1.29–1.76; Table 4) when compared to white men. Factors associated with AI also varied by race/ethnicity. For instance, sex with an injection drug user was independently associated with AI among white men and not Hispanic or African American men. Chlamydia was associated with AI among the overall group of men, with the association being primarily seen among African American men. Specifically, black men who were diagnosed with chlamydia were less likely to report AI (AOR = 0.64; 95% CI: 0.48–0.84).
In this cross-sectional study of a diverse group of clients attending public STD clinics, we found that a substantial minority of clients (10%) reported recent heterosexual AI. The prevalence of AI in this study was lower than that found in the few studies which have examined recent (vs. lifetime) experience with heterosexual anal intercourse.6–8,20 The difference may partly be explained by the way in which AI was ascertained. In this study we assessed AI using face-to-face interviews, which may underestimate the true prevalence of sensitive behaviors.21 Furthermore, the other studies on heterosexual AI were conducted in different populations using restrictive inclusion criteria including substance use, those at high-risk for HIV-infection, or participants enrolled in HIV prevention trials. However, this study is one of the few which examines the prevalence of heterosexual AI among a large and diverse population of clients attending public STD clinics and provides a more representative estimate of the prevalence of AI among this group.
Our findings were similar to other studies, in that STD clinic clients who reported heterosexual AI were more likely to report risky behaviors including exchange of drugs/money for sex, anonymous sex, and substance use.7–9,12 These associations suggest that those who engage in AI also participate in other behaviors that place them at increased risk of acquiring STIs including HIV. However, few studies have examined the association between laboratory confirmed STI infection and recent history of heterosexual AI. While no associations with STIs and AI were found in the overall group, differences were noted by race/ethnicity. For instance, among women we found an association with AI and STIs among Hispanic and white women but not African American women. This may be partly explained by condom use. Unprotected AI was higher among Hispanic and white women (70% reported never using condoms) as compared to African American women (57% reported never using condoms). Likewise, condom use for vaginal intercourse also varied by race/ethnicity, with 35% of African American women reporting unprotected vaginal intercourse as compared to 51% of Hispanic women. Unfortunately condom use was not consistently assessed among all women (>20% missing) and therefore this limits our interpretations on reported condom use.
The difference in STIs and AI by race/ethnicity may also reflect not just individual behavior, but the behavior of one's sexual partner. In particular, we found that white women diagnosed with early syphilis were more than five times as likely to report recent AI. The majority of early syphilis cases in Los Angeles County are diagnosed among men who report sexual contact with men.22 We also found a higher prevalence of early syphilis among male STD clinic clients who reported sexual contact with both men and women (4%), which was substantially higher than that seen among heterosexual men in our clinics (0.4%) or within other comparable STD clinic settings (<1%).23 Furthermore, there is evidence to indicate that men who have sex with both men and women are more likely to practice AI with their female partners.8,20 Therefore, it is possible that white women practicing AI are more likely to have male partners who have sex with both men and women, thus bridging sexual networks, and increasing the likelihood that they are exposed to STIs including HIV.
Among African American men, after accounting for other relevant factors, those diagnosed with Chlamydia were less likely to have reported recent AI as compared to those who did not have chlamydia. The unexpectedly lower prevalence of Chlamydia among black men who reported AI may partly be explained by the age difference across levels of race/ethnicity. Based on our data, we found that African American men who report AI are older compared to both Hispanic and white men who reported AI (mean age = 34, 30, and 27, respectively, P < 0.01). Furthermore, based on the epidemiology of Chlamydia, adolescents and young adults have the highest incidence and prevalence of Chlamydia.24
Among women, those who reported sex with both men and women were more likely to report AI as compared to those who reported only sex with men, which supports the findings from a number of other studies.20,25 However, because of the way in which data were collected, we were unable to conduct a similar analysis for men. Specifically, it is unclear whether anal insertive sex reported by men who had sex with both men and women was practiced with an opposite sex partner. This has important implications in terms of the spread of STIs and HIV in that such “bridging” may enhance the movement of STIs between sexual networks by exposing an individual from a low risk network to a person more likely to have an STI.26,27
Substance use was consistently associated with AI among both genders. Other studies have found associations of heterosexual AI with cocaine, crack, and methamphetamine use; however, this study suggests that other substances including ecstasy, gamma hydroxybutyrate, heroin, and marijuana may also play a role.7,12,28 We also noted an association between engaging in AI and exchange of drugs or money for sex. Although based on our data it is not clear if AI is practiced in exchange for drugs/money or if stimulant use facilitates risky sexual behaviors, this finding emphasizes the role of substance use in heterosexual AI.
Our study has several limitations. Assessment of AI and other sexual risk behaviors was based on self-report and assessed by clinic staff employing face-to-face interview methods. Although this information was collected in the context of health care, interview-based data on socially stigmatized or illicit activities may suffer from reliability and validity issues. During an interview, patients may be reluctant to disclose information regarding sensitive, socially stigmatized, or illegal activities, resulting in response bias and a potential underestimation of these behaviors.21,29,30 Only a small proportion of women reporting AI were tested for anorectal infections (19% tested for rectal Chlamydia; 46% tested for rectal gonorrhea), which underestimates the true prevalence of STIs and limits our ability to assess the association between anal intercourse and rectal STIs. Although the level of underestimation is unknown, it is estimated that among women, 7% to 19% of gonorrhea infections and 21% of Chlamydia infections would be missed without rectal screening.31,32 In terms of external validity, this study was based on participants who attended public STD clinics and therefore may not be generalizable to other populations.
Our results indicate that recent heterosexual AI was reported by a nontrivial proportion of clients seen at public STD clinics. Furthermore, those who reported AI were also more likely to report sexual risk behaviors that place them at high-risk for transmitting and acquiring STIs, and in fact were associated with a concurrent STI. These results have implications for those who provide medical care to clients at STD clinics and highlight the need for rectal screening recommendations for women, specific patient counseling messages related to condom use for AI, the risks associated with unprotected heterosexual AI, substance use, and the increased risk for the transmission or acquisition of HIV and other STIs.
1. Leichliter JS, Chandra A, Liddon N, et al. Prevalence and correlates of heterosexual anal and oral sex in adolescents and adults in the United States. J Infect Dis 2007; 196:1852–1859.
2. 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.
3. Flannery D, Ellingson L, Votaw KS, et al. Anal intercourse and sexual risk factors among college women, 1993–2000. Am J Health Behav 2003; 27:228–234.
4. Baldwin JI, Baldwin JD. Heterosexual anal intercourse: An understudied, high-risk sexual behavior. Arch Sex Behav 2000; 29:357–373.
5. Gorbach PM, Manhart LE, Hess KL, et al. Anal intercourse among young heterosexuals in three sexually transmitted disease clinics in the United States. Sex Transm Dis 2009; 36:193–198.
6. Satterwhite CL, Kamb ML, Metcalf C, et al. Changes in sexual behavior and STD prevalence among heterosexual STD clinic attendees: 1993–1995 versus 1999–2000. Sex Transm Dis 2007; 34:815–819.
7. Koblin BA, Hoover DR, Xu G, et al. Correlates of anal intercourse vary by partner type among substance-using women: Baseline data from the UNITY study. AIDS Behav. In press.
8. Tian LH, Peterman TA, Tao G, et al. Heterosexual anal sex activity in the year after an STD clinic visit. Sex Transm Dis 2008; 35:905–909.
9. Reynolds GL, Latimore AD, Fisher DG. Heterosexual anal sex among female drug users: U.S. national compared to local long beach, California data. AIDS Behav 2008; 12:796–805.
10. Aral SO, Patel DA, Holmes KK, et al. 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.
11. 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.
12. Gross M, Holte SE, Marmor M, et al; The HIVNET Vaccine Preparedness Study 2 Protocol Team. Anal sex among HIV-seronegative women at high risk of HIV exposure. J Acquir Immun Defic Syndr 2000; 24:393–398.
13. Padian N, Marquis L, Francis DP, et al. Male-to-female transmission of human immunodeficiency virus. JAMA 1987; 258:788–790.
14. 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 Immune Defic Syndr Hum Retrovirol 1999; 20:295–299.
15. Skurnick JH, Kennedy CA, Perez G, et al. Behavioral and demographic risk factors for transmission of human immunodeficiency virus type 1 in heterosexual couples: Report from the heterosexual HIV transmission study. Clin Infect Dis 1998; 26:855–864.
16. Guimaraes MD, Munoz A, Boschi-Pinto C, et al; Rio de Janeiro Heterosexual Study Group. HIV infection among female partners of seropositive men in Brazil. Am J Epidemiol 1995; 142:538–547.
17. 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.
18. Liang KY, Zeger SL. Longitudinal data-analysis using generalized linear-models Biometrika 1986; 73:13–22.
19. Zeger SL, Liang KY, Albert PS. Models for longitudinal data—A generalized estimating equation approach. Biometrics 1988; 44:1049–1060.
20. Lescano CM, Houck CD, Brown LK, et al. Correlates of heterosexual anal intercourse among at-risk adolescents and young adults. Am J Public Health 2008; 66:1131–1136.
21. Newman JC, Des JD, Turner CF, et al. The differential effects of face-to-face and computer interview modes. Am J Public Health 2002; 92:294–297.
22. Los Angeles County Department of Public Health. Sexually Transmitted Disease Program. Early Syphilis Surveillance Summary. Los Angeles, CA: Los Angeles County Department of Public Health, 2008:1–39.
23. Warner L, Klausner JD, Rietmeijer CA, et al. Effect of a brief video intervention on incident infection among patients attending sexually transmitted disease clinics. PLoS Med 2008; 5:e135.
24. Centers for Disease Control, Prevention. Sexually Transmitted Disease Surveillance, 2006. Atlanta, GA: U.S. Department of Health and Human Services, 2007.
25. Foxman B, Aral SO, Holmes KK. Heterosexual repertoire is associated with same-sex experience. Sex Transm Dis 1998; 25:232–236.
26. Anderson RM, Holmes KK, Sparling PF, et al. Transmission Dynamics of Sexually Transmitted Infections. Sexually Transmitted Diseases, Vol. 3. New York, NY: McGraw-Hill Companies, Inc, 1999:25–38.
27. Potterat JJ, Muth SQ, Rothenberg RB, et al. Sexual network structure as an indicator of epidemic phase. Sex Transm Infect 2002; 78(suppl 1):i152–i158.
28. 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.
29. Catania JA, Gibson DR, Chitwood DD, et al. Methodological problems in AIDS behavioral research: Influences on measurement error and participation bias in studies of sexual behavior. Psychol Bull 1990; 108:339–362.
30. Fendrich M, Johnson TP, Sudman S, et al. Validity of drug use reporting in a high-risk community sample: A comparison of cocaine and heroin survey reports with hair tests. Am J Epidemiol 1999; 149:955–962.
31. Guerry S, Boudov MR, Higgins CA, et al. Assessing the utility of rectal screening in high risk women: Prevalence of cervical and rectal gonorrhea among females in juvenile detention in Los Angeles County, USA. Paper presented at: International Society for Sexually Transmitted Disease Research; July 29-August 1, 2007; Seattle, WA.
32. Philip SS, Kohn RP, Bernstein JD, et al. Rectal gonorrhea and chlamydia infectins are present in female STD clinic patients without a history of anal sex. Paper presented at: IDSA 46th Annual Meeting; October 25–28, 2008; Washington, DC.