Black gay men run higher risk of HIV infection despite fewer partners

Hyman Scott, University of California, San Francisco, presenting at CROI 2013.
This article is more than 11 years old. Click here for more recent articles on this topic

An analysis of four studies of sexual risk and HIV infection in US gay men, presented at the 20th Conference on Retroviruses and Opportunistic Infections (CROI 2013), has found a 22% higher risk of HIV infection per sexual contact in black gay men that is not explained by other factors such as number of sexual partners, injecting drug use or age.


Young men who have sex with men (MSM) and MSM of colour have the highest HIV incidence rates in the US. Black people form 12% of the US population but have 45% of new HIV diagnoses and, while new HIV diagnoses stayed steady amongst most population groups in the last three years, they increased by 48% in young black gay men.

This is not explained by individual risk behaviour: young gay black men actually have fewer partners and lower rates of recreational drug use than other gay men.

The study

Researchers from the University of California, San Francisco, and the San Francisco Department of Public Health, used data from surveys of gay men that were conducted from 1992 to 2003 to find out if young men and black men had a higher per-contact risk of HIV infection that was not explained by other behavioural or demographic factors.



Receptive anal intercourse refers to the act of being penetrated during anal intercourse. The receptive partner is the ‘bottom’.


Insertive anal intercourse refers to the act of penetration during anal intercourse. The insertive partner is the ‘top’. 

risky behaviour

In HIV, refers to any behaviour or action that increases an individual’s probability of acquiring or transmitting HIV, such as having unprotected sex, having multiple partners or sharing drug injection equipment.

phase III

The third and most definitive stage in the clinical evaluation of a new drug or intervention, typically a randomised control trial with the new intervention compared to an existing therapy or a placebo, in large numbers of participants (typically hundreds or thousands). Trial results are used to evaluate the overall risks and benefits of the drug and provide the information needed for regulatory approval.


How well something works (in a research study). See also ‘effectiveness’.

They looked at four studies:

  • Jumpstart was a CDC study of HIV incidence in high-risk HIV uninfected gay and bisexual men conducted in 1992 to 1994, before combination antiretroviral therapy (ART) was available. 
  • The HIVNET VPS (Vaccine Preparedness Study) was a behavioural survey of gay men in preparation for possible HIV vaccine trials conducted in 1995 to 1997, as ART was becoming available.
  • There were two studies in the ART era: EXPLORE, described in this report, a large behaviour-change trial in gay men, and VAX004, also described in the same report, the first phase III trial of an HIV vaccine, both conducted in 1998 to 2003.

In total there were 10,760 gay men in these studies who paid 42,395 six-monthly trial visits; there were 584 HIV infections in all.

The analysis was restricted to visits where participants reported at least one episode of unprotected receptive anal intercourse, unprotected insertive anal intercourse, protected receptive anal intercourse, unprotected sex with a partner of unknown HIV status, or unprotected sex with more than one HIV-negative partner.

The CDC researchers looked at how, in the four different trials, the number of sexual contacts varied the risk of acquiring HIV.  They then related these to the kind of sex (insertive, receptive protected or receptive unprotected) men reported and whether the partner’s status was HIV positive, HIV negative, or unknown. They also asked about injecting drug use, though in the analysis this had no effect on HIV infections.

Using these figures, they were able to calculate the chance of becoming infected with HIV per sexual contact of a given type and to what degree unprotected sex increased this risk.


The risk of HIV infection per sexual contact with a known HIV-positive partner was one in 137 contacts for unprotected receptive anal sex (0.73%), one in 455 contacts for unprotected insertive sex (0.22%), and one in 1250 contacts for protected receptive sex (0.08%), all with partners known to have HIV.

Interestingly, one can infer from this an 89% prevention efficacy for condom use for receptive anal sex – higher than that seen in another CDC study that compared reported condom use over a six-month period to HIV infections over the same period.

These figures are for the three studies that took place after antiretroviral availability. Figures were little different in the Jumpstart study, with per-contact risks of one in 167, one in 714 and one in 2500 for unprotected receptive and insertive sex and protected receptive sex respectively. These risks were not statistically different from those in the other studies.

Averaged across studies, the risk for unprotected receptive sex with partners of unknown HIV status was one per 204 contacts – not much less risky than with known HIV-positive partners.

Risk varied between studies, with the HIV risk in VAX004 72% higher and in EXPLORE 13% lower than in the VPS study.

Unadjusted for other factors, men under 25 had a 31% higher per-contact risk of being infected, and black men had a 78% higher risk.

This was not because black men had more unprotected sexual contacts. On the contrary, they had considerably fewer – just three for receptive sex and five for insertive sex on average in six months, compared with eleven and twelve respectively for white men.

The average per-contact risk varied with age: for unprotected receptive sex (with an HIV positive partner) it was 1 in 102 for under-25s, 1 in 115 for 25 to 30 year olds, and 1 in 156 for the over-30s.

It varied just as much if not more for ethnicity: it was 1 in 141 for white men (and almost the same for Latino), and 1 in 96 for black men and ‘others’ (which were mainly men of mixed race). However, because of relatively few trial visits and HIV infections in black men, this difference was not actually statistically significant.

Adjusting it by age reduced the increased risk to a 22% higher risk for black men (because they tended to be younger), but an added risk remained.

Unanswered questions

Why the higher risk for black men? This is the unanswered question, and isn’t explained by this study – the findings of which have been observed in other studies too. Black men were in a minority in the studies used, and made fewer study visits, so the figures for them have more uncertainty.

It may be simply because HIV prevalence is higher in US black men so they are generally in a higher-risk environment. Health inequalities also mean that partners are less likely to have undetectable viral loads.

Other factors include network effects, meaning that black men tend to have sex partners from a smaller pool of mainly black partners, and more age-mixing, meaning that black men in some studies have been found to be more likely to have partners who are significantly older than them – and therefore more likely to have HIV.

Other factors may include different degrees of knowledge of partners’ HIV status, and presenter Hyman Scott commented that there was some evidence that black men might not so often be doing other things that reduce HIV risk, such as withdrawal before ejaculation.


Scott H et al. Age and Racial Disparities in Per-contact Risk of HIV Seroconversion among Men Who Have Sex with Men: US. 20th Conference on Retroviruses and Opportunistic Infections, Atlanta, Abstract 91, 2013.

View abstract 91 on the conference website.

A webcast of the session in which this research was presented, North and South: epidemiology and engagement in care, is available on the conference website.