Cherpes, Thomas L. MD*; Meyn, Leslie A. MS†; Hillier, Sharon L. PhD‡
MEDICAL STUDENTS HAVE TRADITIONALLY been taught that herpes simplex virus type 1 (HSV-1) infections most often occur above the waist, whereas the domain of herpes simplex virus type 2 (HSV-2) is below the belt. Especially in young women, however, it appears that these guidelines have become untenable against the numerous reports that have incriminated HSV-1 as an important genital pathogen. In the 1990s, HSV-1 was reported to have replaced HSV-2 as the principle causative agent of genital herpes among young women in Scotland.1 In an investigation of patients from several Swedish sexually transmitted disease (STD) clinics, nearly 65% of primary genital herpes isolates from all adults were the result of HSV-1, and when only women were considered, the virus accounted for approximately 85% of the primary genital HSV infections.2 Likewise, in a Norwegian STD clinic, HSV-1 was responsible for 70% to 90% of the first episodes of genital herpes infections in women less than 21 years of age.3 HSV-1 has also been identified as the most frequent cause of recurrent genital ulcerations among females attending genitourinary medicine clinics in Ireland.4 Most recently, a retrospective analysis of positive genital herpes isolates from a university student health service in the United States found that HSV-1 was more commonly isolated in women than men, and that the proportion of newly diagnosed HSV-1 genital herpes infections had increased from 31% in 1993 to 78% in 2001.5
Our primary objective was to identify the variables associated with the prevalence and acquisition of HSV-1 serum antibodies in a cohort of young, sexually active women. The most frequently identified demographic factors associated with HSV-1 infection have been older age and lower socioeconomic status,6–8 but the specific behaviors associated with HSV-1 infection have not been well delineated. Recently, a study of the general population of England and Wales demonstrated an earlier increase in the prevalence of HSV-1 serum antibodies in females when compared with males of similar age.9 HSV type-specific serologic enzyme immunoassays, like the one used in that investigation, could not discriminate between the oral or genital acquisition of HSV-1, but an earlier increase in disease frequency among women is characteristic of a sexually transmitted pathogen. Similarly, the strongest predictor of HSV-1 infection among individuals from a London STD clinic was an earlier age at first intercourse; another behavior variable typical of a sexually transmitted infection.10
Materials and Methods
A total of 1207 study participants were recruited from three Pittsburgh-area health clinics for an investigation designed to identify risk factors associated with the acquisition of vaginal group B Streptococcus.11 The 3 sites of enrollment included a low-income primary health care clinic, a university-based student health clinic, and the local county health department’s STD clinic. Written informed consent was obtained, and approval for the subsequent use of the collected data was obtained from the Institutional Review Board of the Magee-Women’s Hospital of Pittsburgh and the study participants. Nonpregnant women 18 to 30 years of age who sought medical care that required a gynecologic examination were eligible for participation. Exclusion criteria included the presence of vaginal bleeding, the current use of systemic antimicrobials, or the use of any intravaginal products in the 24 hours before enrollment.
From 1998 through 2000, demographic and behavioral interview data as well as serum for HSV type-specific antibody testing were obtained from all enrolled women. After the initial visit, the women were requested to return for 3 follow-up visits, at 4-month intervals, and the sera collected at each visit were stored at −70°C before HSV antibody testing. Five hundred sixteen of the 1089 women who returned for at least 1 follow-up visit were HSV-1-seronegative at enrollment, whereas 737 women were HSV-2-seronegative. These represented the subsets of women considered for the longitudinal analysis of risk factors associated with HSV-1 or HSV-2 antibody acquisition.
The association between the presence of bacterial vaginosis and the acquisition of HSV-2 detected in this same cohort of women has been previously published.12
Serum specimens were tested for type-specific HSV-1 and HSV-2 antibodies using 2 commercially available enzyme immunoassays (ELISAs) from Focus Technologies (Cypress, CA). Both assays are approved for use on adults by the U.S. Food and Drug Administration and were used according to the manufacturer’s instructions. In head-to-head comparison with Western blot analysis, the sensitivity and specificity of the Focus HSV-1 ELISA are 98% and 94%, respectively, whereas those of the Focus HSV-2 ELISA are 98% and 95%, respectively.13,14 As detailed in the package insert, index values ≥1.10 were considered positive, whereas values <0.90 were classified negative. Serum samples for which the obtained results were equivocal (index values between 0.90 and 1.09 inclusive) were retested. Specimens again identified as equivocal after retesting were excluded from the analysis.
The enrollment data were analyzed with SPSS statistical software, release 10.1.4 (SPSS, Inc., Chicago, IL). Seven hundred forty-seven of the 1207 study participants (61.9%) described themselves as white, 412 (34.1%) as black, and 48 (4.0%) as Hispanic, Asian, Native American, or multiethnic. The latter 48 women were combined into 1 racial category referred to as “other” for analysis. Univariate associations of baseline characteristics with HSV-1 seroprevalence were tested with Pearson’s chi-square test, a chi-square test for linear trend, or Fisher exact test, where appropriate. All statistical tests were evaluated at the 0.05 level of significance. Variables with P values less than 0.1 were considered for inclusion in a multivariable logistic regression model. Enrollment site was not included in the model as a result of its collinearity with all other variables (race, age, education, number of sex partners, and history of a STD). Models were developed with use of forward stepwise regression based on the likelihood ratio test statistic. Variables were retained in the model if the Wald chi-square test statistic had a P value of 0.05 or less.
The longitudinal data were analyzed using Stata statistical software, release 8.0 (Stata Corp., College Station, TX). Acquisition of HSV-1 or HSV-2 was defined as a change in the ELISA index value from <0.9 (negative) at enrollment to ≥1.1 (positive) at the 4-, 8-, or 12-month follow-up visits. Log rank tests for the equality of survivor functions were used to evaluate the risk factors associated with the acquisition of HSV-1. Comparisons between the acquisition rates of HSV-1 and HSV-2 antibodies were made using a 2-sample exact test for binomial probabilities.
HSV-1 seroprevalence in the combined study population was 47%, but the prevalence varied widely by site of enrollment. Although fewer than 30% of the women enrolled from the university student health clinic were HSV-1-seropositive, half or more of the women enrolled from both the primary care and STD clinics had serum HSV-1 antibodies (Table 1). Univariate analysis of the enrollment data also revealed statistically significant associations between the presence of HSV-1 antibodies and several demographic characteristics: black race, older age, and fewer years of education, and multivariable logistic regression analyses of the cross-sectional data also demonstrated these same variables to be among the independent predictors of HSV-1 seroprevalence (Table 1).
HSV-1 seroprevalence was higher among those women who reported at least 5 lifetime male sex partners in both the univariate and multivariate analysis. Women who reported a sexual history that included receptive oral sex were somewhat more likely to be HSV-1-seropositive, but this association did not reach statistical significance. In addition, no statistically significant differences in the HSV-1 seroprevalence were detected among women who reported either a history of intercourse with an uncircumcised male partner or a history of sex with another woman.
The seroprevalence of HSV-1 among women who had previously been told by a physician or nurse that they had an oral herpes infection was 85% (23 of 27), compared with a 46% (539 of 1180) seroprevalence among those who denied such a history, and receiving a diagnosis of oral herpes from a healthcare worker was also among the independent risk factors for HSV-1 seroprevalence by multivariate analysis (Table 2). However, the sensitivity for a history of an oral herpes infection diagnosed by a physician or nurse among the HSV-1-seropositive women in our investigation was only 4% (23 of 562) (Table 2). Common reproductive tract infections, including trichomoniasis, chlamydia, gonorrhea, and pelvic inflammatory disease were all associated with HSV-1 seropositivity by univariate analysis, but these relationships did not remain statistically significant in the multivariate analyses. HSV-1 seroprevalence among women who either denied or reported a history of a diagnosis of genital herpes by a healthcare worker was approximately the same (47% vs. 45%) (Table 2).
The 516 HSV-1 seronegative women identified at enrollment returned for 1833 follow-up visits, from which 485 woman-years of follow up were accumulated. During follow up, there were 29 women who acquired antibodies to HSV-1. This yielded an overall HSV-1 acquisition rate of 6.0 cases per 100 woman-years of follow up. Twelve of the seroconversions occurred among women enrolled from the university student health clinic, 7 were from the primary care clinic and 10 were among those initially enrolled from the STD clinic (Table 3).
Further characteristics of the women who acquired antibodies to HSV-1 are shown in Table 3. In the initial analysis, no significant risk factors for HSV-1 acquisition were found. For example, HSV-1 acquisition was not associated with enrollment site, race, age, or education. Likewise, no significant associations among the study participants were detected between the acquisition of HSV-1 and the frequency of vaginal intercourse, the presence of a new male sex partner, or higher numbers of male sex partners in the 4 months before a follow-up visit. Women whose partners had used condoms were as likely to acquire HSV-1 as those who partners did not (5.8 vs. 6.3 cases per 100 woman-years; P = 0.8). This is in contrast to a recent study of HSV-2-serodiscordant couples in which male condom use offered significant protection against HSV-2 acquisition.15 Although the rate of HSV-1 acquisition was higher among those women who did not have an antecedent HSV-2 infection compared with those women who were HSV-2-seronegative (6.8 vs. 2.9 cases per 100 woman-years), this relationship did not achieve statistical significance (P = 0.2).
In our investigation, the rate of HSV-1 acquisition was slightly higher than the acquisition rate for HSV-2 (6.0 cases vs. 4.6 cases per 100 woman-years of follow up, respectively; P = 0.3). This differs from a previous HSV acquisitional study, which reported a similar rate of new HSV-2 infection (5.1 cases per 100 person-years), but a much lower rate of new HSV-1 infection (1.6 cases per 100 person-years).16 However, that study differed from ours in that it measured the rates of HSV acquisition in discordant couples rather than women alone. Because HSV-2 is typically a genitally acquired infection, whereas HSV-1 is acquired through both oropharyngeal and genital modes of transmission, we further examined the data for differences in acquisition rates for either virus by the specific sexual activities reported by the women in the 4 months before seroconversion (Table 4). Not unexpectedly, there were no HSV-2 acquisitions among the women who were sexually abstinent. More surprising was the observation that women who reported vaginal intercourse in the preceding 4 months were significantly more likely to acquire HSV-1 than were those women who were sexually abstinent (6.8 vs. 1.2 cases per 100 woman-years of follow up; P = 0.05) (Table 4).
Among the HSV-1- and HSV-2-susceptible cohorts of women who had vaginal intercourse, the rates of HSV-1 and HSV-2 acquisition were similar (6.8 vs. 5.7 cases per 100 woman-years of follow up, respectively; P = 0.5). Women who had receptive oral sex without vaginal intercourse were found to be much more likely to acquire HSV-1 than were sexually abstinent women (9.8 vs. 1.2 cases per 100 woman-years of follow up; P = 0.04) (Table 4). In comparison, there were no HSV-2 acquisitions among the women who received cunnilingus but denied vaginal intercourse in the 4 months before HSV-2 seroconversion or among those women who reported sexually abstinence (Table 4). These results suggest that although HSV-2 is more likely to be acquired through genital transmission, both vaginal intercourse and receptive oral sex may be important for HSV-1 acquisition among sexually active women.
In this study of 18- to 30-year-old women, rates of HSV-1 and HSV-2 acquisition were similar and appeared to be associated with sexual activity. Receptive oral sex was identified as a risk factor for HSV-1, but not HSV-2, acquisition. Our findings are supported by previous cross-sectional studies. Results from the Swedish investigation of individuals who presented at STD clinics with primary genital herpes infections demonstrated that significant correlates of HSV-1 infection included a history of oral–genital sex and a history of herpes labialis in the partner of the study participant.2 Similarly, results from a STD clinic in the United States reported that a history of receptive oral sex was significantly associated with an increased likelihood for isolation of HSV-1 from a positive genital HSV culture,17 although in an investigation of university students in Germany and Spain, the only behavioral factor found associated with HSV-1 seroprevalence was a history of sexual intercourse.18
Recurrent herpes labialis is the most frequent clinical manifestation of HSV-1 infection; for example, it is present in approximately 38% of young adults in the United States.19 HSV-1 is also known to be present in salivary excretions in the absence of any clinically obvious oral lesion.20 Apparently, asymptomatic shedding of HSV-1 in oral secretions is relatively common, having been reported in as many as 9% of adults.21 Therefore, among HSV-1-seronegative women, receptive oral sex from a partner with oral herpes should be considered an activity that can result in transmission of the virus.
Our results are also reconciled with proposed explanations for the emergence of HSV-1 as a significant genital pathogen. At enrollment, we detected HSV-1 antibodies in only 38.4% (Table 1) of the 18- to 20-year-old women. This low frequency of infection is consistent with previous reports that indicate a reduction in the age-specific HSV-1 seroprevalence in the United States.22,23 The falling rate of childhood HSV-1 infections seen in the United States, and other developed countries, has resulted in a burgeoning proportion of young adults who are susceptible to either oral or genital HSV-1 infection. This HSV-1-naive population of young adults is participating in oral sex. Although it is difficult to determine if sexual behavior patterns have significantly changed, the lifetime prevalence of receptive oral sex among sexually active women is 75%.24 In our cohort of 18- to 30-year-old women, more than 90% of the study participants reported a history of receptive oral sex. Although it is possible that the increased significance of HSV-1 as a STD is a result of changes in viral pathogenicity, it seems likely that the decreasing incidence of HSV-1 acquisition in childhood, and the likelihood these younger cohorts will participate in oral sex, accounts for the increased recognition of HSV-1 as a sexually transmitted infection.
There are several implications behind the observation that HSV-1 has become responsible for an increased percentage of genital herpes infections. First, seroprevalence studies such as the National Health and Nutrition Examination Survey (NHANES) will significantly underestimate the prevalence of genital herpes in a population if only participants with type-specific antibodies to HSV-2 are considered. Second, because neonatal herpes infections can result from the acquisition of an HSV-1 genital infection during pregnancy, awareness of the risks, especially in the third trimester, of cunnilingus among HSV-1-discordant couples needs to be increased. Although HSV-1 is only currently responsible for 30% of the cases of neonatal herpes in the United States, it has been identified as the causative agent in half of the cases in the United Kingdom and in nearly 75% of the cases in The Netherlands.25,26 Third, although recent evidence suggests that symptomatic penile ulcerations caused by HSV-1 are relatively rare,27 asymptomatic shedding and transmission of the virus during intercourse likely does occur. Furthermore, although recurrent penile lesions are more likely to be caused by HSV-2, recurrent rectal/perirectal lesions in men are more likely to be caused by HSV-1.27 Therefore, it is possible that the use of male condoms may be less likely to be effective to prevent HSV-1 transmission during sexual intercourse. Finally, although the development of a genital herpes vaccine is an area of current investigation, most of this research has focused on vaccines directed against HSV-2-specific targets. Because our investigation suggests that HSV-2 infection does not confer protection against the acquisition of HSV-1 (Table 2), an optimally effective genital herpes vaccine may also need to include HSV-1-specific targets that reduce susceptibility for the acquisition of HSV-1.
The identification of cunnilingus as a risk factor for HSV-1 acquisition in women is consistent with the changing epidemiology of genital herpes infections documented in many developed countries. Although our investigation is one of the largest studies of HSV-1 acquisition to date, the analyses may be limited by the small number of new HSV-1 infections and woman years of follow up among those study participants who reported receptive oral sex without vaginal intercourse (Table 4). Another limitation is our dependence on type-specific serologic assays to identify women who acquired HSV-1 infections. Because these assays cannot distinguish between an oropharyngeal or genital mode of HSV-1 acquisition, the relative contributions of each cannot be determined. Furthermore, it is not possible to exclude kissing as the behavioral activity responsible for HSV-1 acquisition. A related limitation is the absence of information pertaining to the presence of genital or oral HSV-1 infections among the sexual partners and social contacts of the women in our investigation. Our study also did not adequately assess the proportion of times with which male condoms were used during sexual intercourse, and this limitation may explain our failure to detect a benefit of condom use among women who acquired HSV-1. Despite these limitations, this study has documented a low incidence of HSV-1 seroconversion among young sexually abstinent women and an increased risk for HSV-1 acquisition among those women having vaginal intercourse or receptive oral sex.
1. Scoular A, Leask B, Carrington D. Changing trends in genital herpes due to herpes simplex virus type 1 in Glasgow, 1985–1988 [Letter]. Genitourin Med 1990; 66:226.
2. Löwhagen G, Tunbäck P, Andersson K, et al. First episodes of genital herpes in a Swedish STD population: A study of epidemiology and transmission by the use of herpes simplex virus (HSV) typing and specific serology. Sex Transm Infect 2000; 76:179–182.
3. Nilsen A, Myrmel H. Changing trends in genital herpes simplex virus infection in Bergen, Norway. Acta Obstet Gynecol Scand 2000; 79:693–696.
4. Coyle PV, O’Neill HJ, Wyatt DE, et al. Emergence of herpes simplex type 1 as the main cause of recurrent genital ulcerative disease in women in Northern Ireland. J Clin Virol 2002; 27:22–29.
5. Roberts CM, Pfister JR, Spear SJ. Increasing proportion of herpes simplex virus type 1 as a cause of genital herpes infection in college students. Sex Transm Dis 2003; 30:797–800.
6. Lamey P, Hyland PL. Changing epidemiology of herpes simplex virus type 1 infections. Herpes 1999; 6:20–24.
7. Nahmias A, Roizman B. Infection with herpes simplex viruses 1 and 2 (third of three parts). N Engl J Med 1973; 289:781–789.
8. Malkin J, Morand P, Malvy D, et al. Seroprevalence of HSV-1 and HSV-2 infection in the general French population. Sex Transm Infect 2002; 78:201–203.
9. Vyse AJ, Gay MJ, Slomka MJ, et al. The burden of infection with HSV-1 and HSV-2 in England and Wales: Implications for the changing epidemiology of genital herpes. Sex Transm Infect 2000; 76:183–187.
10. Cowan FM, Copas A, Johnson AM, et al. Herpes simplex virus type 1 infection: a sexually transmitted infection of adolescence? Sex Transm Infect 2002; 78:346–348.
11. Meyn LA, Moore DL, Hillier SL, et al. Association of sexual activity with colonization and vaginal acquisition of group B Streptococcus in nonpregnant women. Am J Epidemiol 2002; 155:949–957.
12. Cherpes TC, Meyn LA, Krohn MA, et al. Association between acquisition of herpes simples virus type 2 in women and bacterial vaginosis. Clin Infect Dis 2003; 37:319–325.
13. Ribes JA, Smith A, Hayes M, et al. Comparative performance of herpes simplex virus type 1-specific serologic assays from MRL and Meridian Diagnostics. J Clin Microbiol 2002; 40:1071–1072.
14. Ribes JA, Hayes M, Smith A, et al. Comparative performance of herpes simplex virus type 2-specific serologic assays from Meridian Diagnostics and MRL Diagnostics. J Clin Microbiol 2001; 39:3740–3742.
15. Wald A, Langenberg AGM, Link K, et al. Effect of condoms on reducing the transmission of herpes simplex virus type 2 from men to women. JAMA 2001; 285:3100–3106.
16. Langenberg AG, Corey L, Ashley RL, et al. A prospective study of new infections with herpes simplex virus type 1 and type 2. N Engl J Med 1999; 341:1432–1438.
17. Lafferty WE, Downey L, Celum C, et al. Herpes simplex virus type 1 as a cause of genital herpes: Impact on surveillance and prevention. J Infect Dis 2000; 181:1454–1457.
18. Stock C, Guillén-Grima F, Hermosa de Mendoza J, et al. Risk factors of herpes simplex type 1 (HSV-1) infection and lifestyle factors associated with HSV-1 manifestations. Eur J Epidemiol 2001; 17:885–890.
19. Embril JA, Stephens RG, Manuel FR. Prevalence of recurrent herpes labialis and aphthous ulcers among young adults in six continents. Can Med Assoc J 1975; 113:627–630.
20. Douglas RG, Couch RB. A prospective study of chronic herpes simplex virus infection and recurrent herpes labialis in humans. J Immunol 1970; 104:289–295.
21. Overall JC Jr. Dermatologic viral diseases. In: Gallasso GJ, Merigan TC, Buchanen RA, eds. Antiviral Agents and Viral Diseases of Man, 2nd ed. New York: Raven Press, 1984:247–312.
22. Nahmias AJ, Lee FK, Beckman-Nahmias S. Seroepidemiological and sociological patterns of herpes simplex virus infection in the world. Scand J Infect Dis (Suppl) 1990; 69:19–36.
23. Xu F, Schillinger JA, Sternberg MR, et al. Seroprevalence and coinfection with herpes simplex virus type 1 and type 2 in the United States, 1988–1994. J Infect Dis 2002; 185:1019–1024.
24. Sadock VA. Normal human sexuality and sexual dysfunctions. In: Kaplan and Sadock’s Comprehensive Textbook of Psychiatry, 7th ed. Philadelphia: Lippincott Williams & Wilkins, 2000: 1577–1631.
25. Tookey P, Peckham CS. Neonatal herpes simplex virus infection in the British Isles. Paediatr Perinat Epidemiol 1996; 10:432–442.
26. Gaytant MA, Steegers EAP, Cromvoirt PLM, et al. The incidence of neonatal herpes in The Netherlands. Ned Tijdschr Geneeskd 2000; 144:1832–1836.
27. Solomon L, Cannon MJ, Reyes M, et al. Epidemiology of recurrent genital herpes simplex virus types 1 and 2. Sex Transm Infect 2003; 79:456–459.