HIV spreads in those slutty phases: new evidence from a UK study

In a fascinating new paper published in the wonderful, open-access Public Library of Science, Fraser Lewis and colleagues shed a little more light on how HIV spreads. They show that, at least among gay men in the UK, HIV propels itself in bursts, exploding quickly through a whole cluster of people and then tailing off.

They also show that a quarter of all new infections occured within the first six months after a person was infected, and close to half in the first year. These data are hugely important. There’s a big push in the United States and some other countries at the moment to focus HIV prevention on people who are already infected. This makes some sense; only people who have the virus can pass it on, so if they could be persuaded to use clean needles and condoms the problem would be solved, right? Yes but. The but here is that if you take this approach, you can only focus on known positives. Most people in most countries aren’t diagnosed until they’ve been infected for many years. And as this new study confirms, by that time they’ve already infected most of the people they are going to infect.

We’re back to something I’ve posted on before: the peaks and troughs of our sex lives, aka our slutty phases. Most infection gets passed on soon after you’re infected. And very often you get infected when you’re climbing one of your sex life’s peaks, when you are out there having a good time, often with several partners in just a few months. (Aside — how come the troughs always seem to last so much longer than the peaks?). Lewis isn’t actually looking at behaviour, he and his colleagues base their analysis on genetic similarities and differences in the virus itself, but the conclusion is inevitable.

In my mind, the clear implication is that we need to focus HIV prevention efforts in physical places and in communities where we are likely to intercept people in their slutty phases — in clubs and bars, on internet dating sites and in chat rooms. Waiting until someone is diagnosed as infected before cranking up prevention is just not good enough.

By the way, if you’re intimidated by the nerd-speak in the paper itself (or even by its title: Episodic Sexual Transmission of HIV Revealed by Molecular Phylodynamics), just read the Editors’ summary. I cannot commend PLoS too highly for these wonderful, accurate but extraordinarily user-friendly summaries. They represent a huge step forward in the public communication of science.

This post was published on 19/03/08 in Science.

Send this post to a friend Send this post to a friend


You can follow the comments on this post via this RSS feed.

Tags: , , , , , .

  1. Comment by Chris Green, 20/03/08, 01:38:


    First, you may wish to correct the link of this PLoS article – it should be

    I think your assumption that supporters of Positive Prevention advocate “Waiting until someone is diagnosed as infected before cranking up prevention…” is a little facile. The article that many consider the grandfather of this movement (The Serostatus Approach to Fighting the HIV Epidemic: Prevention Strategies for Infected Individuals – SAFE, by Robert Janssen et al, including your ‘epi-hero’ Kevin De Cock, Am J Public Health. 2001;91:1019–1024, ) proposes these Essential Components:

    “The 5 SAFE steps are as follows: (1) Increase the number of HIV-infected persons who know their serostatus. (2) Increase the use of health care and preventive services. (3) Increase high-quality care and treatment. (4) Increase adherence to therapy by individuals with HIV. (5) Increase the number of individuals with HIV who adopt and sustain HIV–STD risk reduction behavior.”

    Recently infected people DO often experience symptoms of acute infection. These symptoms do sometimes cause them to go to the doctor. But rarely do they cause the doctor to consider HIV infection as the reason, and then ask the questions which might encourage the person to get tested.

    Similarly the International HIV/AIDS Alliance publication “Positive Prevention” notes that “People test for HIV when it makes sense for them to do so. Therefore it is important to work towards increasing the number of reasons for people to come forward for testing. Availability of ARV treatment is a major reason.”

    I have frequently suggested that we should determine the reasons why people decide to get tested for HIV. In this way, we use are eforts to promote testing (limited as they are) to strengthen these reasons.

    I’m sure you would accept that prevention efforts require multiple “focuses”; positive prevention, in particular as proposed by the SAFE approach, must be one important element of these efforts.


  2. Comment by Lee Rudolph, 20/03/08, 03:16:

    “They also show that a quarter of all new infections occured within the first six months after a person was infected, and close to half in the first year.” I have read that there are various viruses that induce quite marked changes in the behavior of their hosts (for instance, making some sort of buggish creature climb to the tops of grassblades, where passing birds can more easily find them and eat them) in ways that favor the transmission of the virus. Is it possible that HIV, in the first 6 to 12 months of residence in a new human host, induces an increase in that host’s sluttishness? And if that were somehow found (at the moment, I can’t quite imagine the experimental protocol) to be likely true, what would a good public-health response be?

  3. Comment by elizabeth, 20/03/08, 10:11:

    Chris: of course I’m not suggesting that we abandon the SAFE approach. And yes, I would advocate that we increase people’s awareness of the symptoms of primary infection — doctors as well as people whose behaviours put them at risk (as I’ve said in earlier posts). I’m just concerned that with the increasing focus on access to treatment, what little is left of the prevention agenda is getting skewed more and more towards those known to be infected. Yes, we need to provide prevention services for positives. But we also need to renew efforts to increase safe behaviour in areas where the majority of new infections are likely to be taking place.

    Lee: You’d have to control for the ordinary cycles of sluttish and non-sluttish behaviour in people without the virus, I suppose. Can I be on your IRB?

  4. Comment by Chris Green, 20/03/08, 12:02:

    Lee, there’s no doubt that a person is most infectious in the period before seroconversion, when the viral load its probably at its all-time peak. That’s why the virus spreads so quickly among groups of injecting drug users (IDU). In fact, although I have difficulty in fully understanding what’s new in the article, it does seem to me that the risk behaviour of the MSM surveyed is similar to that of IDU, except perhaps they didn’t have sex together three times a day, so the period would be longer. Of course, once the group is all infected, no more infections take place as long as no-one breaks out of the group. And when he does, he may by then have a lower viral load, and so a lower infectiousness, so it takes a little longer for the next cluster to form.

Comments are closed at this time.