It sounds like there are at least two archtypes for sexual networks, which for lack of better terms, I’d call vertical and horizontal. In non-hierarchical sexual networks of concurrent partnerships, it wouldn’t matter much where HIV entered as it would quickly burn through as everyone is in a sense a ‘node’ – there are fewer deadends in the tracing the contacts. In the gonorrhea network example, however, there were clear hierarchies of contacts, with ‘core spreaders’ at the primary hubs and sexual contacts branching off into fewer numbers of subsequent contacts.

You raise an excellent point and I have to agree that under the conditions you imply with concurrent partnerships, male circumcision will be most effective when adopted broadly.

Hi Ben,
i couldn’t figure out how to create an account on the blog to comment on your post about Male circumcision as policy. There is one statement you make in the post, that i would caution – You say “Circumcision, like any STI control strategy, works when core spreaders at the nodes of the sexual networks are treated.” While this is accurate based on many of the assumptions driving a LOT of interventions, it doesn’t reflect the reality of sexual networks in SSA (and may not anywhere, but that’s beyond the scope of this particular point). The thing is, few, if any, sexual networks actually have any such “hubs”. Recent empirical evidence demonstrates this (see a forthcoming piece by Helleringer and Kohler in AIDS), and a handful of people have been
suggesting it for a while (see especially Martina Morris’s work). This particular point doesn’t have much bearing on the “end-game” of the suggestion you make in this post, but is something that has been proliferated without much empirical backing for a while. Epidemics don’t need hubs, so interventions need to be broad based.

Jimi Adams, PhD
Sociology
OSU