Abstract
Sir: In a recent report, McQuillan et al. 1 present correlates of prevalent HIV infection gleaned from the National Health and Nutrition Examination Surveys (NHANES) conducted from 1999 to 2006. The NHANES samples are representative of the non-institutionalized population of the United States aged 18–49 years and include nearly 12,000 participants. Multivariate analyses, which controlled for history of homosexual contact and injection drug use, revealed that in non-Hispanic black men, lifetime number of sex partners was not positively associated with prevalent HIV infection. The special attention to black men's HIV prevalence is merited not only because they comprise the single most severely affected demographic group (by sex and race) in the United States, but also because their HIV infection base rate is sufficiently high to allow stable statistical associations.
It is surprising that this lack of association was neither discussed, nor its crucial implication elaborated by McQuillan and her colleagues. This finding is consistent with other studies showing no association between number of sex partners and prevalent HIV infection, especially studies that adequately controlled for sex of participants, heterosexual anal intercourse, male homosexual activity and non-sexual blood exposures. 2–5
The substantial independent association between prevalent HSV-2 infection and prevalent HIV infection in the NHANES data also is unlikely to represent sexual HIV transmission. In randomized trials, HSV-2 suppressive therapy reduced genital ulcers substantially but led to a non-significant tendency toward increasing the risk of acquiring HIV. 6,7 HIV infection often precedes HSV-2 infection. 8,9 and prior HIV infection is associated with incident HSV-2 infection 10,11 perhaps due to immunosuppression.
In sum, McQuillan and colleagues' report adds to the epidemiological and microbiological 12–14 evidence that penile–vaginal intercourse in reasonably healthy adults – as opposed to anal intercourse or blood exposures – is not a significant mode of HIV transmission.