Transmission of Herpes Simplex Virus Types 1 and 2 in a Prospective Cohort of HIV-Negative Gay Men: The Health in Men Study
- Fengyi Jin1,
- Garrett P. Prestage1,
- Limin Mao2,
- Susan C. Kippax2,
- Catherine M. Pell6,
- Basil Donovan1,3,
- David J. Templeton1,
- Janette Taylor5,
- Adrian Mindel4,
- John M. Kaldor1 and
- Andrew E. Grulich1
- National Centres in
- 1HIV Epidemiology and Clinical Research and
- 2HIV Social Research, University of New South Wales,
- 3Sydney Sexual Health Centre, Sydney Hospital, and
- 4Sexually Transmitted Infections Research Centre, Westmead Hospital and University of Sydney, Sydney,
- 5Westmead Millennium Institute, Centre for Virus Research, Westmead Hospital, Westmead, and
- 6Taylor Square Private Clinic, Darlinghurst, New South Wales, Australia
- Reprints or correspondence: Mr. Fengyi Jin, National Centre in HIV Epidemiology and Clinical Research, Level 2, 376 Victoria St., Darlinghurst, NSW 2010, Australia (jjin{at}nchecr.unsw.edu.au)
Abstract
BackgroundDespite increasing reports of herpes simplex virus (HSV) type 1 (HSV-1)–associated anogenital herpes, there are very limited data comparing the seroepidemiological profile of and risk factors for HSV-1 and HSV type 2 (HSV-2) infection
MethodsSexual behaviors were examined as risk factors for prevalent and incident HSV-1 and HSV-2 infections in a community-based cohort of 1427 HIV-negative gay men in Australia
ResultsThe prevalence of HSV-1 and HSV-2 at baseline was 75% and 23%, respectively. The rate of prevalent infection with HSV-1, as well as the rate of prevalent infection with HSV-2, was much lower in individuals <25 years of age, and each type of infection was associated with a higher number of both male and female sex partners. The median duration of follow-up of the cohort was 2.0 years. Among participants who were susceptible to infection, the incidence rates for HSV-1 and HSV-2 infection were 5.58 and 1.45 cases per 100 person-years, respectively. In multivariate analysis, incident infection with HSV-1 was significantly associated with younger age (P=.027) and reports of frequent insertive oral sex with casual partners (hazard ratio, 3.91 [95% confidence interval, 1.23–12.44]; P=.021). Incident infection with HSV-2 was significantly associated with a variety of anal sex practices with casual partners
ConclusionsBoth HSV-1 and HSV-2 were commonly sexually transmitted, and there were more HSV-1 than HSV-2 seroconversions. Public-health strategies targeted against anogenital herpes increasingly need to take into account the importance of HSV-1 infection
There are 2 types of herpes simplex virus (HSV): HSV-1 and HSV-2. HSV-1 typically causes oropharyngeal infection, and transmission occurs primarily through personal contact during childhood, although it can also be transmitted sexually. HSV-2 is mainly associated with genital disease, and transmission is usually sexual [1]. Both viruses can cause either anogenital or oropharyngeal infection [2]
Recently, a change in the relative contribution of HSV-1 and HSV-2 to anogenital herpes has been noted. HSV-1 has become an increasingly common cause of clinical anogenital herpes in developed countries, including Australia [3–7]. It has been postulated that lower rates of HSV-1 acquisition during childhood and increasing rates of orogenital sex practices may explain this finding [8], but there are few longitudinal data to confirm this. Epidemiological studies of HSV-2 have shown this virus to be more common in homosexually active men [8–10], and they have reported an independent association with HIV acquisition in this population [11, 12]. Given the increasing importance of HSV-1–related anogenital herpes, we examined the seroepidemiological profile of and risk factors for both HSV-1 and HSV-2 infections in a longitudinal study of gay men in Sydney, Australia
Participants and Methods
ParticipantsStudy participants were men in the Health in Men (HIM) cohort study who were recruited from June 2001 through December 2004. The methods used in the HIM study have been described in detail elsewhere [13]. In brief, men eligible for participation in the HIM study met the following criteria: (1) they reported having sex with other men within the 5 years before enrollment in the study, (2) they lived in Sydney or participated regularly in the gay community of Sydney, and (3) they tested negative for HIV at baseline. Participants were recruited from a range of community-based settings. Signed, informed consent was obtained from all participants. Ethics approval was granted by the University of New South Wales
Data collectionAll men who were eligible to participate in the study underwent annual face-to-face interviews, with telephone interviews conducted every 6 months between these visits. Areas covered in the interview included lifetime sexual contacts, detailed sexual behaviors occurring in the past 6 months, self-report of episodes of genital herpes occurring in the past 12 months, and demographic factors. Blood samples were obtained at each annual face-to-face interview
At each interview, detailed quantitative data on the number of episodes of insertive and receptive unprotected anal intercourse (UAI) occurring in the past 6 months were collected separately for regular sex partners (hereafter referred to as “regular partners”) and for casual sex partners (hereafter referred to as “casual partners”), according to the HIV status (negative, positive, or unknown) of these partners, and, for receptive intercourse, according to whether or not ejaculation occurred. In addition, semiquantitative data (data on whether practices occurred never, occasionally, or often) were also collected for the following sex practices: insertive and receptive oral sex, anal fingering, anilingus (rimming), fisting, use of dildos, intercrural intercourse (thigh sex), wet and dry kissing, sensuous touching, mutual masturbation, urolangia (water sports), sexual sadomasochism, bondage and discipline, and group sex. These data were also recorded separately for regular and casual partners
Laboratory studiesTesting was performed in 2005, with the use of serum samples that were obtained from June 2001 through June 2005 and were stored at −20°C until testing. Serum samples were tested for antibody to HSV by use of the Enzygnost anti-HSV IgG ELISA (Behring) for the detection of total antibody to HSV. If the serum samples had a positive test result, they were tested using the following algorithm. First, they were tested for antibody to HSV-2 by use of an indirect ELISA (HerpeSelect 2 ELISA IgG; Focus Technologies). The manufacturer indicates that an index value of >1.10 is evidence of the presence of antibody to HSV-2. However, there is evidence that use of this cutoff value yields a high rate of false-positive results. To overcome this issue, a cutoff value of 3.5 was used to denote HSV-2–seropositive serum samples [14]. Serum samples for which the HerpeSelect 2 ELISA (index values, 0.9–3.5) provided an equivocal result underwent further testing by HSV-2 Western blot analysis [15]. Depending on the HSV-2 serostatus associated with these serum samples, HSV-1 serostatus was determined. If a serum sample was determined to have a positive result by use of the Enzygnost anti-HSV IgG ELISA and was found to be HSV-2 antibody negative by the HerpeSelect 2 ELISA, the serum sample was presumed to be HSV-1 antibody positive. If a serum sample was determined to have a positive result by use of the Enzygnost anti-HSV IgG ELISA and was found to be HSV-2 antibody positive by the HerpeSelect 2 ELISA, the serum sample was tested for the presence of antibody to HSV-1 by use of the HerpeSelect 1 ELISA IgG (Focus Technologies) and HSV-1 Western blot analysis. This testing algorithm has been used and validated elsewhere by our group [16]
Sequentially obtained serum samples were tested as follows. All consenting participants were tested for HSV-1 and HSV-2 at baseline. If they were found to be seronegative to either virus, their last available blood sample drawn between July 2004 and June 2005 was retested for that same virus(es). For those participants who had experienced seroconversion, all interim samples were also tested to determine the time of seroconversion
Previous or current hepatitis B virus (HBV) infection at baseline was determined by use of an HBV core (HBcAb) antibody test (AxSYM CORE; Abbott Diagnostics). The screening test for syphilis was an EIA (ICE Syphilis; Murex Biotech). Positive results of EIAs were confirmed using the Treponema pallidum particle-agglutination assay and the florescent treponemal antibody absorption test
Statistical analysisStatistical analyses were performed using Stata software (version 8.2; Stata). The exact binomial method was used to calculate 95% confidence intervals (CIs) for prevalence and incidence values. For the calculation of incidence, among individuals who initially were found to be seronegative, the date of infection was assumed to be the midpoint between the time when the last negative test result was noted and the time when the first positive test result was noted. Total person-years (PYs) were calculated as the time from study entry to either the estimated date of infection or the time when the last negative test result was noted for participants who remained seronegative
Logistic regression was used to identify the risk factors associated with prevalent infections, and Cox regression was used to identify risk factors associated with incident infections. Crude and adjusted analyses were performed to identify factors associated with HSV-1 and HSV-2 infections. For ordinal variables, such as age, number of sex partners, and frequencies of specific sexual behaviors, P values for trend were reported. For nonordered variables, tests for homogeneity were performed. Odds ratios (ORs), hazard ratios (HRs), and their corresponding 95% CIs were calculated for these associations. Multivariate logistic regression and Cox regression models were developed to determine the risk factors that were independently associated with HSV seroprevalence and incidence. Variables for which the P value was <.10 in univariate analyses were considered in the multivariate analyses. For incident infections, the age groups of the study participants were recategorized into 4 groups (<25 years, 25–34 years, 35–44 years, and >44 years), because of the limited number of participants who had seroconversion. Sexual behaviors were examined separately according to regular and casual partners, and the results are presented separately in the text (for regular partners) and in the text and tables (for casual partners). Analyses for regular and casual partners were not exclusive (i.e., participants who reported both regular and casual partners were included in both regular and casual partner analyses)
Results
From June 2001 through December 2004, a total of 1427 participants were enrolled in the study. Participants were recruited from a large variety of sources, including outdoor gay community events (55.3% of participants), referrals from other participants (12.5%), gay venues (6.5%), previous studies (6.6%), gay organizations (4.6%), the Internet (4.1%), and the gay press plus other sources (6.6%). Only 3.8% of participants were recruited from clinics. The median participant age at enrollment was 35 years (range, 18–75 years). The majority of participants (95.2%) self-identified as gay or homosexual
HSV-1 and HSV-2 prevalenceA total of 1390 participants (97.4%) consented to HSV testing, and stored serum samples were available for 1371 participants (96.1%). At baseline, 1031 participants (75.2% [95% CI, 72.8%–77.5%]) were HSV-1 seropositive, 319 (23.3% [95% CI. 21.1%–25.6%]) were HSV-2 seropositive, and 258 (18.8% [95% CI, 16.8%–21.0%]) were both HSV-1 and HSV-2 seropositive
In univariate analysis, both prevalent HSV-1 and prevalent HSV-2 infections were significantly associated with increasing age, a higher lifetime number of reported male and female sex partners, and serological evidence of previous HBV infection (anti-HBcAb) at baseline. In addition, HSV-2–seropositive status was also associated with the participants rating anal intercourse as an important part of their sexual behavior, as well as with a positive result of serological testing for syphilis at baseline (table 1)
Univariate analysis of risk factors for prevalent herpes simplex virus type 1 (HSV-1) and type 2 (HSV-2) infection in the Health in Men cohort study
In multivariate analyses, several factors were significantly associated with both prevalent HSV-1 and prevalent HSV-2 infections, including a higher lifetime number of reported male and female sex partners, previous HBV infection, and a lower education level (table 2). In addition, prevalent HSV-2 infection was significantly associated with increasing age and a preference for being the receptive partner when engaging in anal intercourse
Multivariate analysis of risk factors for prevalent herpes simplex virus type 1 (HSV-1) and type 2 (HSV-2) infection in the Health in Men cohort
Incidence of HSV-1 and HSV-2 infection among participants susceptible to infection at baselineBy June 2005, a total of 1239 participants (86.8%) had completed at least 1 follow-fup face-to-face interview. The total follow-up was 2686 PYs (median follow-up, 2.0 years). Of 340 men who were found to be HSV-1 seronegative at baseline, 33 had seroconversion to HSV-1, for an incidence of 5.58 cases of HSV-1 infection/100 PYs (95% CI, 3.87–7.74 cases) among participants who were susceptible to infection. Of 1052 men who were found to be HSV-2 seronegative at baseline, 28 had seroconversion to HSV-2, for an incidence of 1.45 cases of HSV-2 infection/100 PYs (95% CI, 0.97–2.09 cases) among participants who were susceptible to infection. There were no participants who had both HSV-1 and HSV-2 seroconversion
In each 6-month interview interval, an average of 70.7% of participants reported having sexual contacts with regular partners, and an average of 75.4% of participants reported having sexual contacts with casual partners. In univariate analysis, incident HSV-1 infection was significantly associated with younger participant age, and the rate of such infection peaked at 11.5 cases/100 PYs among participants <25 years of age (table 3). The following sexual practices with regular partners were significantly associated with incident HSV-1 infection: a higher number of regular partners in the past 6 months (Ptrend=.009), having UAI with sex partners who had an unknown HIV status or sex partners who were HIV positive (HR, 3.17 [95% CI, 1.06–9.44]; P=.039), and frequent oroanal (Ptrend=.033) and orogenital (Ptrend=.033) sex practices. In multivariate analysis, a higher number of regular partners (Ptrend=.015) remained significant, and the association with frequent oroanal sex practices with regular partners was of borderline significance (Ptrend=.069). The following sex practices with casual partners were significantly associated with incident HSV-1 infection (table 4): a higher number of casual partners in the past 6 months and insertive oral sex with ejaculation (table 4). In multivariate analysis, younger age (Ptrend=.027) and frequent reporting of insertive oral sex with ejaculation with casual partners (HR, 3.91 [95% CI, 1.23–12.44]; P=.021) were associated with a higher risk of incident HSV-1 infection
Association between study participant age and incidence of herpes simplex virus type 1 (HSV-1) and type 2 (HSV-2) infection in the Health in Men cohort study
Univariate analysis of sexual behaviors as risk factors for incident herpes simplex virus type 1 (HSV-1) and type 2 (HSV-2) infections among men in the Health in Men cohort study who reported having casual sex partners
In contrast to incident HSV-1 infection, there were no HSV-2 seroconversions in participants <25 years of age, and men <35 years of age were at a decreased risk of incident HSV-2 infection, compared with older men (table 3). None of the reported sexual behaviors with regular partners was associated with an increased risk of incident HSV-2 infection. However, a variety of anal sex behaviors with casual partners were significantly associated with incident HSV-2 infection, including receptive UAI with ejaculation, fingering, oroanal practices, and use of a dildo (table 4). In multivariate analysis, frequent insertive fingering (Ptrend=.025) and receiving a dildo (P = .030) with casual partners were significantly associated with incident HSV-2 infection
Association between HSV-1 and HSV-2 infectionThe presence of the one type of HSV antibody at baseline was associated with the presence of the other type of HSV antibody (OR, 1.53 [95% CI, 1.12–2.08]). Thirty of the 33 HSV-1 seroconversions occurred in men who were previously found to be HSV-2 negative, and 6 of the 28 HSV-2 seroconversions occurred in men who were previously found to be HSV-1 negative. Previous HSV-2 antibody–positive status did not protect against subsequent acquisition of HSV-1 (HR, 0.42 [95% CI, 0.13–1.39]), and previous HSV-1–positive status did not protect against subsequent acquisition of HSV-2 infection (HR, 1.18 [95% CI, 0.48–2.90])
Genital herpes outbreaksAt baseline, 0.4% of participants who tested negative for both HSV-1 and HSV-2 reported having had an episode of genital herpes in the past 12 months. Among participants who tested positive for HSV-1 only, 3.2% reported a recent episode of genital herpes. For HSV-2, 23.7% of participants who tested positive for HSV-2 only reported a recent episode of genital herpes, compared with 10.2% of those who tested positive for both HSV-1 and HSV-2 (P=.005)
Discussion
The present study confirms that HSV-1 infection is now a common sexually transmitted infection in homosexually active men in developed countries. Among study participants who, at baseline, were susceptible to acquiring either HSV-1 or HSV-2 infections, the incidence of HSV-1 infection was nearly 4 times greater than the incidence of HSV-2 infection, and there were more seroconversions to HSV-1 than to HSV-2. Analysis of risk factors suggested differing modalities of sexual transmission for the 2 HSV types. Incident HSV-1 infection was associated with younger age and with oroanal and orogenital sexual practices. On the other hand, incident HSV-2 infection was more common in older men and was associated with anal sex practices. At baseline, men who tested positive only for HSV-2 were more likely to report a recent episode of genital herpes than were men who tested positive for both HSV-1 and HSV-2
As a prospective cohort study, the HIM study was able to provide data on both the prevalence and incidence of HSV infection. There is a considerable body of previous research on the prevalence of HSV-1 and HSV-2 infections in homosexually active men. The observed seroprevalence of HSV-1 (75%) was almost identical to that reported for homosexual men in a clinic-based survey in Australia [17]. Although 75% was only slightly greater than the seroprevalence of 71% found in a population-based survey of men in the general population, the prevalence of HSV-1 infection in HIM study participants <45 years of age was substantially greater than that in the general population. For older men, rates were more similar to those in the general population [16]. The fact that, for young, but not older, homosexual men, the prevalence of HSV-1 was higher than the prevalence of HSV-1 in the general population supports the hypothesis that HSV-1 has only recently become commonly sexually transmitted in this population. Elsewhere, the prevalence of HSV-1 in homosexual men has been reported to be 55%–77% in a small number of studies in Europe, the United States, and Japan [8]. The striking variation in HSV-1 prevalence that occurs with age makes direct comparison between studies difficult. However, the studies are consistent in showing that HSV-1 is highly prevalent [8, 10]
In contrast, the HSV-2 seroprevalence (23%) observed in the HIM study is nearly 3 times as high as that noted for men in the general population in Australia [16], and it was similar to that reported for HIV-negative homosexual men attending clinics in Melbourne, Australia [17]. Studies in other developed countries also confirm that homosexually active men are at increased risk of HSV-2 infection [8, 18–23]
Our findings regarding the risk factors for prevalent HSV-1 and HSV-2 infection are remarkably consistent with those of previous studies. Increasing age, a higher number of lifetime sex partners, a history of STDs, and a lower education level have all been reported as risk factors for both prevalent HSV-1 and prevalent HSV-2 infection [16, 17, 19–36]. In addition, studies of homosexual men have also reported that receptive UAI is a risk factor for HSV-2 [19, 20]
Somewhat surprisingly, in this cohort of gay men, the lifetime number of female sex partners was associated with both HSV-1 and HSV-2 positivity. A similar association—but only with HSV-2—was recently reported in a study of homosexual men in the United States [20]. For HSV-2 and, to a lesser extent, for HSV-1, seroprevalence in the general population is higher in women than in men [8, 10, 16, 29, 34, 37]. The association with heterosexual contact could be related to the late establishment of homosexual contact in some homosexually active men, as well as previous sexual activity with women. Consistent with this hypothesis, men who reported a higher lifetime number of female sex partners in the HIM study reported an older age at first homosexual anal intercourse (P<.01; data not shown)
Data on the incidence of HSV among homosexual men are very limited. Compared with the rate of HSV-2 infection, there was a very high rate of HSV-1 infection among susceptible individuals (5.58 cases/100 PYs), and rates of infection were >10% per year in individuals <25 years of age. Rates of HSV-2 infection were somewhat lower (1.45 cases/100 PYs) and were highest in individuals 35–44 years of age. The only previous study of HSV incidence in a community-based sample of homosexual men was a much smaller US study that reported an incidence of HSV-1 infection of 4.3 cases/100 PYs and an incidence of HSV-2 infection of 1.0 cases/100 PYs [38]. A clinic-based study of individuals who had undergone repeated testing for HIV in the United States reported a higher incidence of HSV-2 infection (3.1 cases/100 PYs) but did not report data on the incidence of HSV-1 infection [19]
Our analysis of risk factors for incident HSV infection in homosexual men includes much larger numbers of individuals who experienced seroconversion than did previous studies [19, 38]. Orogenital and oroanal practices were significant predictors of incident HSV-1 infection in univariate analysis, and insertive orogenital sex remained significant in the multivariate model. The increasing occurrence of orogenital sex has been suspected to be the reason for the increasingly common occurrence of HSV-1–related genital herpes [5, 6, 8, 39], and our data provide longitudinal evidence to support this hypothesis. Another factor that was strongly suggestive of sexual transmission of HSV-1 was the markedly higher prevalence of HSV-1 infection in individuals 25–34 years of age (76%), compared with that in individuals <25 years of age (54%). For incident HSV-2 infection, anal sex practices were more closely associated with seroconversion. In the multivariate model, the association with anal fingering and use of dildos remained significant
The present study had the strength of being community based, and participants can be considered to be broadly representative of HIV-negative men who are involved in the gay community in Sydney. In contrast, studies based on clinic samples may overrepresent the number of men considered to be at high risk for HSV infection, and their results may be less generalizable. The behavioral data sought in the HIM study relied on the self-report of participants, and no attempt was made to verify those reports. Even though the present study included the largest number of individuals with HSV-1 and HSV-2 seroconversion of any prospective study involving homosexual men, it had relatively limited power to detect risk factors for incident HSV infection. Future testing of this cohort is likely to provide further useful information
In recent years, HSV-1–associated genital herpes has increased in frequency, and, in some settings, it has overtaken HSV-2 as the most common cause of initial genital herpes, particularly in young persons [6]. Although it is usually accepted that HSV-1 is less likely to cause recurrent disease of the genitals, at least 1 study has reported that HSV-1 may have become the most common cause of recurrent genital ulcers [40]. In the sexually transmissible infection and HIV clinic setting, HSV-1 has been reported to cause only a minority of recurrent lesions, but HSV-1 infection is more likely in persons with recurrent perianal lesions [41]. The serological data from the HIM study complement and extend the data from these clinical series
Almost one-half of young gay men <25 years of age in the HIM study were susceptible to HSV-1 infection, and seroconversion rates were >10% per year. As previously suspected, orogenital and oroanal practices were associated with the risk of HSV-1 infection. Sexual transmission of HSV-1 is likely to become increasingly common in developed countries, driven by increasing rates of orogenital and oroanal contact [42, 43], in a population that is more likely to be susceptible because of lower rates of childhood HSV-1 infection. Public health strategies aiming to decrease the influence of anogenital herpes on health, including its influence on HIV transmission, increasingly need to consider both HSV types
Footnotes
-
↵Financial support: The National Centre in HIV Epidemiology and Clinical Research and the National Centre in HIV Social Research are funded by the Australian Government Department of Health and Ageing. The project was funded by the Australian Government Department of Health and Ageing (Canberra) and the New South Wales Health Department (Sydney). The Health in Men Cohort Study was funded by the National Institutes of Health (NIH), a component of the US Department of Health and Human Services (NIH/National Institute of Allergy and Infectious Diseases/Division of AIDS: HVDDT award N01-AI-05395) and the National Health and Medical Research Council in Australia (grant 400944). Herpes testing was funded by GlaxoSmithKline
Potential conflicts of interest: D.J.T. and A.E.G. have received financial support for conference attendance and honorariums for educational activities from GlaxoSmithKline. All other authors: no conflicts
- Received March 26, 2006.
- Accepted April 28, 2006.
- © 2006 by the Infectious Diseases Society of America










