Secondary Logo

Journal Logo

Original Studies

Role of Sexual Behavior in the Acquisition of Asymptomatic Epstein-Barr Virus Infection

A Longitudinal Study

Woodman, Ciaran B. J. MD*; Collins, Stuart I. BSc*; Vavrusova, Nicol BSc; Rao, Ankit MB, BcH BAO*; Middeldorp, Jaap M. PhD; Kolar, Zdenek MD; Kumari, Angela BSc*; Nelson, Paul PhD§; Young, Lawrence S. DSc*; Murray, Paul G. PhD*

Author Information
The Pediatric Infectious Disease Journal: June 2005 - Volume 24 - Issue 6 - p 498-502
doi: 10.1097/01.inf.0000164709.40358.b6
  • Free


The Epstein-Barr virus (EBV) is a common herpesvirus that is implicated in a range of malignant diseases; however, the natural history of EBV infection is poorly defined. It has been hypothesized that infectious mononucleosis is usually acquired from a transfer of saliva and that in young adults this is more likely after the onset of sexual activity.1,2 Although this hypothesis has been widely accepted, only limited evidence can be adduced to support an association between the acquisition of EBV infection in young adults and their sexual behavior. Some cross-sectional studies have reported that sexually experienced women are more likely to have antibodies to EBV3,4 than those who have yet to have intercourse; others have been unable to confirm this association.5 An association with the number of sexual partners and early age of first intercourse has also been reported for patients with infectious mononucleosis but not for those with asymptomatic infection.3,4 However, such cross-sectional observations have only a limited ability to reveal an association between infection and sexual behavior, because they cannot distinguish asymptomatic infections acquired in childhood from those acquired after the onset of sexual activity. When the time of onset of infection cannot be determined, all sexual exposures must be included in the analysis, although some might only have occurred after the infection was acquired and others so long before that they can have had no causal significance. However, restriction of the analysis to include only new sexual partners acquired shortly before the detection of infection can also fail to reveal a true association with sexual behavior if infectiousness is intermittent and if the probability of sexual transmission is low.6 Only a longitudinal study design can describe the risk of infection in relation to changes in sexual behavior. However, to date such studies have provided little evidence that asymptomatic infection is related either to “deep kissing” or to sexual intercourse, 2 activities that are clearly associated: in one of these studies, there are only 3 subjects in the exposed group, and in the other, exposure status is not adequately defined.3,7 We report the prevalence and subsequent incidence of EBV infection in a cohort of sexually active young women and explore the social and sexual determinants of incident infections.


The study population is drawn from a cohort of 2011 young women enrolled in a longitudinal study of risk factors for early cervical neoplasia.8 In brief, between 1988 and 1992, women between 15 and 19 years of age who were likely to be available for long term follow up because of their continuing need for contraceptive supplies were recruited from a single family planning clinic; they were seen at intervals of ∼6 months, when their social and sexual risk factor profile was updated, and cytologic and serologic samples were taken and stored for subsequent analysis. A total of 1023 women contributed serologic samples to the prevalence study; all women who tested negative for antibodies to EBV in their first sample and who had 1 or more subsequent samples contributed to the longitudinal study (n = 45). Cervical samples were available for 14 of the 22 women who seroconverted during follow-up; these samples were used to describe the temporal relationship between the detection of EBV DNA at this site and seroconversion.

Laboratory Methods

All serum samples were analyzed by enzyme-linked immunosorbent assay (ELISA) for serologic reactivity against both viral capsid antigen (VCA) and Epstein-Barr nuclear antigen 1. Selected samples were analyzed by immunofluorescence (IF) to confirm the ELISA results or to determine the serologic status of those cases, which could not be resolved as positive or negative by ELISA.


For assessment of general EBV serostatus, a synthetic peptide-based Combiplate ELISA assay was used that measures IgG reactivity to a combination of immunodominant epitopes of Epstein-Barr nuclear antigen 1 (BKRF1) and VCA-p18 (BFRF3) according to the manufacturer (Cyto-Barr BV, Bergen, the Netherlands) and as described elsewhere.9,10 Study samples and positive and negative control samples were diluted 1/100 in phosphate-buffered saline (PBS)-Tween buffer containing bovine serum albumin and Triton X (PBS, pH 7.6, with 0.05% Tween 20, 3% bovine serum albumin and 0.1% Triton X-100). Diluted serum (100 μL) was incubated for 1 hour at 37°C in Combiplate wells. After a washing in PBS-Tween buffer (PBS with 0.05% Tween 20) wells were incubated in horseradish peroxidase-conjugated secondary mouse antihuman IgGγ (diluted 1/600 in PBS containing 1% BSA) for 1 hour at 37°C. After a further wash in PBS/Tween buffer, horseradish peroxidase reactivity was visualized with 100 μL of 3,3′5,5′-tetramethylbenzidine substrate system for 15 minutes in the dark. The reaction was stopped by the addition of 2 mol/L sulfuric acid, and optical density was read at 450 nm. Samples were recorded as positive if the optical density of the test sample was greater than the mean of 3 different known EBV-positive serum samples. Samples were recorded as negative if they were below the mean of 3 known EBV-negative samples. Any samples not meeting these criteria were tested by IF (see below). In addition, 25% of all samples identified as either positive or negative by ELISA were retested by IF. To verify the prior seronegative status of seroconverters, VCA-IgM reactivity was determined in a subset of these women with acommercial IgM-capture ELISA (bioMérieux, Boxtel, the Netherlands), which also incorporates the VCA-p18 combipeptide.


Cells from the lytically infected P3HR1 cell line and the latently infected Raji line were spotted onto glass microscope slides and fixed in acetone. After rehydration of slides in heat-inactivated normal goat serum, 10 μL of each serum sample diluted in PBS containing 20% heat-inactivated normal goat serum (1/10, 1/20 and 1/40) was applied to cell spots. After 1 hour of incubation and 2 washes in PBS, pH 7.6, 10 μL of goat anti-human IgG conjugated to fluorescein isothiocyanate (dilution, 1/100 in PBS) were applied. Slides were washed twice in PBS, mounted in a medium containing 1,4-diazobicyclo[2,2,2]octane and recorded as either positive or negative by evaluation under fluorescence microscopy.

EBV DNA Detection.

DNA was extracted from cervical cytology samples using the DNeasy Tissue Kit (Qiagen). EBV DNA was quantified with primers that amplify a fragment of the EBV DNA polymerase gene.11 Amplification of the human β2-microglobulin was used to determine the input cellular DNA in each sample; all samples included in this analysis tested positive for β2-microglobulin. All standards and test samples were analyzed in duplicate (10, 5 and 2 copy number standards were run in triplicate). Each experiment included DNA samples prepared from EBV-positive and EBV-negative controls, as well as water-only (no template) controls. Samples were considered negative if the Ct values exceeded 40 cycles.

Statistical Methods

A case-control analysis, nested within the cohort of 45 women for whom the first EBV sample tested was EBV-negative and who had further follow-up, was undertaken. Cases were women who became EBV-positive during follow-up, with the date of seroconversion defined as the midpoint between the date of the first EBV-positive and the date of the immediately preceding EBV-negative sample. Controls were women who remained EBV-negative throughout follow-up and also cases before seroconversion.12 Each case was matched with 4 controls according to duration of coital experience, measured in continuous time, that is, the exact interval between date of first intercourse and date of seroconversion or the corresponding point in time during follow-up for the matched control; at a given point in time, a future case was eligible as a control only if the samples forming the interval containing that point in time were EBV-negative. This matching method is similar to the “risk-set” concept of Cox proportional hazards regression and meant that controls could appear in a given matched set only once, but could appear in more than one matched set. Conditional logistic regression models were constructed, and confidence intervals for odds ratios were calculated with the use of parameter estimates and standard errors; statistical comparisons were made with the likelihood ratio test.13

To determine whether the acquisition of a new sexual partner is a risk factor for infection with EBV, it is necessary to identify a plausible exposure window (a period of time before seroconversion during which it is possible for the acquisition of a new partner to be a causal factor in that seroconversion); partners acquired before the start, or after the end, of this exposure window are explicitly excluded from the measurement of risk. We examined the risk of EBV infection associated with acquiring a new partner during intervals of varying duration before seroconversion to determine which interval yielded the maximum risk. The minimum time from acquisition to seroconversion was assumed to be 1 month; therefore all exposure windows terminated at the point in time 1 month before seroconversion.


The social and sexual behavioral characteristics of the study populations at study entry are listed in Table 1. Nine hundred seventy-eight (95.6%) women tested positive for antibodies to EBV in their first sample. Of 45 women who tested negative for EBV, 22 seroconverted during follow-up. The median time to seroconversion was 25 months (range, 1–60 months), and the median age at seroconversion was 18 years (range, 16–21 years). The risk of seroconversion increased as the total number of partners increased, but fell with increasing age at seroconversion and was unrelated to age at first intercourse, age of most recent partner or social class of father (Table 2). The risk of infection associated with the acquisition of a new sexual partner was greatest when that partner had been acquired between 1 and 25 months before the seroconversion date, after which it declined (Table 3).

Distribution of Social and Sexual Behavioral Characteristics at Study Entry for the Full Study Population (n = 1023) and for the Women Contributing to the Longitudinal Study (n = 45)
Univariate Analysis of the Relationship Between Study Variables and Seroconversion to EBV
Univariate Analysis of the Relationship Between the Acquisition of a New Sexual Partner During an Exposure Window and the Risk of Infection with EBV

Of 14 seroconverters tested with the IgG ELISA, an IgM response was detected in 8 at the same time as IgG was first detected. All but 1 sample that tested negative by IgG ELISA also tested negative for IgM antibodies to EBV VCA; the remaining sample, which was taken at the visit immediately preceding the first detection of an IgG response, revealed a IgM antibody response.

Cytologic samples from 14 women were tested for the presence of EBV DNA sequences; all tested positive for β2-microglobulin. The median number of samples available for testing was 6 (range, 1–8). Nine women tested positive for EBV DNA at 1 or more visits: 4 tested positive only once; 3 tested positive twice, and 2 tested positive at 3 or more visits (Table 4). In 1 woman, EBV DNA was first detected at the visit immediately preceding the first detection of EBV antibodies (subject 2); in 5 women, at the same time as antibodies were first detected (subjects 1, 3, 6, 7 and 8); in 2, at the visit immediately after the first detection of antibodies (subjects 4 and 5); and in 1 at the second visit after the first detection of antibodies (subject 9). Three women (subjects 2, 4 and 6) had very high EBV DNA loads per cell. Four women were still EBV DNA-positive at the end-of-follow-up (subjects 2, 3, 4 and 6).

Detection of EBV DNA in Cervical Samples Taken From Women Who Seroconverted


The prevalence of antibodies to EBV in this cohort was similar to that reported in a recent population based survey.14 Despite the small sample size, this is the first study to show an association between sexual behavior and the acquisition of an incident asymptomatic infection. Our findings suggest that the acquisition of infection is rarely contemporaneous with the first exposure to an infected partner and that considerable time may have to elapse before a woman becomes infected. This delay may reflect either the intermittent infectiousness of a partner or a low probability of virus transmission after any single episode of sexual intimacy. However, any measure of risk will be imprecise in the absence of information on the infective status of sexual partners For example, the attenuation of risk observed after 25 months may merely reflect the unavoidable inclusion in the analysis of uninfected partners. Although infection in childhood has been associated with parental social class, this does not appear to be a determinant of infection in young adults.15 The fall in risk of seroconversion with increasing age is consistent with the observation that in some people antibodies to EBV cannot be detected, although it seems inconceivable that they have never been exposed to the virus.16 Although our longitudinal design provides the most persuasive evidence to date of an association with sexual behavior, multiple sexual practices are often conflated within the same individual, and we cannot as yet distinguish osculatory and venereal routes of transmission. Although a venereal route of transmission has been suggested, based on the finding of infectious virus in the female genital tract, the detection of EBV DNA in cervical material in this study may simply be a consequence of infection in trafficking lymphocytes and does not necessarily indicate viral replication at this site.17,18 However, the surprisingly high EBV copy number in cervical samples taken from 3 subjects in this series is more consistent with the shedding of cell-free virus, and the continued detection of virus for a prolonged period makes a venereal route of transmission for another herpesvirus more probable.


1. Hoaagland RJ. The transmission of infectious mononucleosis. Am J Med Sci. 1955;229:262–272.
2. Evans AS. Infectious mononucleosis in University of Wisconsin students: report of a five year investigation. Am J Hyg. 1960;71:342–362.
3. Andersson-Ellstrom A, Svennerholm B, Forssman L. Prevalence of antibodies to herpes simplex virus types 1 and 2, Epstein-Barr virus and cytomegalovirus in teenage girls. Scand J Infect Dis. 1995;27:315–318.
4. Crawford DH, Swerdlow AJ, Higgins C, et al. Sexual history and Epstein-Barr virus infection. J Infect Dis. 2002;186:731–736.
5. Andersson-Ellstrom A, Bergstrom T, Svennerholm B, Milsom I. Epstein-Barr virus DNA in the uterine cervix of teenage girls. Acta Obstet Gynecol Scand. 1997;76:779–783.
6. Niederman JC, Miller G, Pearson HA, Pagano JS, Dowaliby JM. Infectious mononucleosis, Epstein-Barr virus shedding in saliva and oropharynx. N Engl J Med. 1976;294:1355–1359.
7. Dan R, Chang RS. A prospective study of primary Epstein-Barr virus infection among university students in Hong Kong. Am J Trop Med Hyg. 1990;42:380–385.
8. Woodman CB, Collins S, Winter H, et al. Natural history of cervical human papillomavirus infection in young women: a longitudinal cohort study. Lancet. 2001;357:1831–1836.
9. van Grunsven W, Spaan J, Middeldorp J. Localisation and diagnostic application of immunodominant domains of the BFRF-3 encoded Epstein-Barr virus capsid protein. J Infect Dis. 1994;170:13–19.
10. Meij P, Vervoort MB, Bloemena E, et al. Antibody responses to Epstein-Barr virus-encoded latent membrane protein-1 (LMP1) and expression of LMP1 in juvenile Hodgkin's disease. J Med Virol. 2002;68:370–377.
11. Junying J, Herrmann K, Davies G, et al. Absence of Epstein-Barr virus DNA in the tumor cells of European hepatocellular carcinoma. Virology. 2003;15:306:236–243.
12. Rothman KJ, Greenland S. Case-control studies. In: Rothman KJ, Greenland S, eds. Modern Epidemiology. 2nd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 1998:93–114.
13. Breslow NE, Day NE. IARC Scientific Publications No. 32. Statistical Methods in Cancer Research. vol 1: The Analysis of Case-Control Studies. Lyon, France: International Agency for Research on Cancer; 1980.
14. Morris M, Edmunds WJ, Hesketh LM, et al. Sero-epidemiological patterns of Epstein-Barr and herpes simplex (HSV-1 and HSV-2) viruses in England and Wales. J Med Virol. 2002;67:522–527.
15. Crowcroft NS, Vyse A, Brown DW, Strachan DP. Epidemiology of Epstein-Barr virus infection in pre-adolescent children: application of a new salivary method in Edinburgh, Scotland. J Epidemiol Community Health. 1998;52:101–104.
16. Wagner HJ, Hornef M, Teichert HM, Kirchner H. Sex difference in the serostatus of adults to the Epstein-Barr virus. Immunobiology. 1994;190:424–429.
17. Portnoy J, Ahronheim GA, Ghibu F, Clecner B, Joncas JH. Recovery of Epstein-Barr virus from genital ulcers. N Engl J Med. 1984;311:966–968.
18. Sixbey JW, Lemon SM, Pagano JS. A second site for Epstein-Barr shedding: the uterine cervix. Lancet. 1986;2:1122–1124.
© 2005 Lippincott Williams & Wilkins, Inc.