METHODS

Between November 1990 and December 1991, 869 consecutive GUM clinic patients attending a representative sample of clinics were invited to participate in the study, as were 1494 consecutive blood donors between February and April 1992. All participants had a blood sample taken for antibodies to HSV and self completed a questionnaire, which collected information on demographic and behavioural characteristics as well as on symptoms/history of herpes infection. Results on HSV-2 seroprevalence and symptoms of herpes infection have been published elsewhere.^7,8

Blood samples were tested for herpes simplex virus antibody at University of Washington by western blot ⁹ which is sensitive and specific for identifying patients with past herpes simplex virus type 1 and herpes simplex virus type 2 or co-infected with both agents.

Statistical analysis was undertaken using stata 6.0. The unadjusted odds of HSV-1 were calculated for sociodemographic and behavioural factors. The adjusted odds were calculated adjusting for all other factors. This complete case analysis was based on 759 (STI clinic attenders), and 1144 (blood donors) participants respectively. In order to examine the extent to which sexual behaviour is associated with HSV-1 acquisition, the calculation of adjusted odds of HSV-1 only included individuals who had ever had sex. In order to separate the effects of age and age at first sexual intercourse, a subset of the data was reanalysed (individuals >25 years who had had sex before age 25). These results are not presented here, but this analysis did not substantially change the results.

RESULTS

The response rate was 98% among eligible participants at both sites. The demographic and behavioural characteristics of the populations are shown in table 1. Sixty four per cent of clinic attenders were single, 84.6% were white and 76.9% were in social class I, II, or III. Thirty five per cent of male and 51% of female donors were single and overall 94% were white. Information on occupation was not collected from blood donors.

The prevalence of antibody to HSV-1 among clinic attenders was 60.4% (95% CI 57.0 to 63.7) and among donors was 46.1% (95% CI 43.5 to 48.7). The demographic and behavioural factors associated with presence of antibody to HSV-1 are shown in table 2. On univariate analysis, the prevalence of antibodies to HSV-1 increased with age in both populations, HSV-1 was more common in black GUM clinic attenders than white and was less common among GUM attenders of higher socioeconomic status. Married blood donors were more likely to be infected than single.

In GUM clinic attenders, the adjusted odds of HSV-1 indicated that HSV-1 was less common among heterosexual men than women and homosexual men, became more common with increasing age and was more common among black people and those of lower socioeconomic status. Among blood donors, only increasing age and being single rather than married were independently associated with HSV-1 infection.

In relation to the behavioural factors, it is striking that early age at first intercourse is much more strongly associated with presence of HSV-1, than lifetime partners in both groups. Among GUM clinic attenders someone who was aged 20 was over 60% less likely to have HSV-1 than someone who had sex for the first time aged 15. Although the association was less marked in blood donors, those aged 20 at first intercourse were 36% less likely to have HSV-1 than those who had sex for the first time aged ≤15.

DISCUSSION

Population based studies of sexual behaviour confirm that the age of first sexual intercourse in Britain has decreased over the past 30 years, although this has likely stabilised over the past decade. ¹⁰ Data from other studies confirm that childhood acquisition of HSV-1 (presumably non-sexual) is declining and that a majority of young people are HSV-1 naive when they start to become sexually active.³ Several centres in the United Kingdom report that primary genital HSV-1 is now the most common cause of genital herpes in young women.⁴

In this study we demonstrate that in the United Kingdom HSV-1 behaves similarly to some of the viral STIs, such as HSV-2 and HIV. It is more common among women and homosexual men than heterosexual men and among GUM clinic attenders than blood donors. Age of first intercourse is the strongest predictor of infection. However, serological evidence of HSV-1 antibody provides no information about likely site of sexually acquired HSV-1. Study participants were asked about past diagnosis and symptoms of both oro-labial and genital herpes. These data have been presented elsewhere ⁸ but in brief indicate that around 16% of GUM attenders with HSV-1 antibody alone had symptoms suggestive of genital herpes but that this rate was much lower in blood donors (3%). Sexual transmission of HSV-1 can result in oro-labial herpes (through kissing) or genital herpes (through either oro-genital or possibly genital-genital contact). However, the fact that more than 60% of genital infections in young women in some centres are due to HSV-1 means that at least some of these infections are likely to be genitally acquired.⁴

These data have two implications for young people. It appears that people who start having sex at an early age are particularly vulnerable to infection with HSV-1, more likely as a result of their sexual behaviour than because of any biological predisposition. Of note, because HSV-1 is transmitted through both social and sexual contact, individuals who have never had sex but are infected with HSV-1 can potentially transmit genital herpes to their sexual partners. Secondly, because transmission studies have not shown that HSV-1 infection is protective of acquisition of infection with HSV-2,^11,12 an individual who acquires genital herpes due to HSV-1 early in life could potentially acquire infection with HSV-2 subsequently, which is more likely to become recurrent.