Background A frequently used HIV risk reduction method among men who have sex with men (MSM) is serosorting, which can be defined as a restriction of sexual partnerships to those who are of the same HIV serostatus. This partner selection strategy has been shown to reduce HIV transmission in epidemiological studies and mathematical models, but may lead to acquisition of other sexually transmitted infections (STI) including syphilis. We sought to evaluate the impact of HIV serosorting on syphilis prevalence, and to assess whether serosorting could account for an observed rise in syphilis incidence among MSM in San Francisco.

Methods A deterministic SIRS (susceptible-infectious-resistant-susceptible) model of syphilis transmission among HIV-negative and HIV-positive MSM was developed; model input parameters were based on epidemiological data from San Francisco between 1998 and 2004. The primary outcome was the impact of HIV serosorting on syphilis prevalence; we further evaluated the influence of HIV prevalence and average number of sexual partnerships on this serosorting effect.

Results Simulations showed that for base-case conditions, HIV serosorting increases syphilis transmission among HIV-positive and also among HIV-negative MSM so that syphilis could become endemic like in San Francisco. Only under very specific circumstances with high levels of serosorting among HIV-negative men can serosorting decrease syphilis prevalence. The size of the impact of serosorting on syphilis prevalence depends on HIV prevalence and partnership number.

Conclusions Our mathematical model adds evidence to the conclusion from an earlier ecological study suggesting that serosorting of HIV-negative and HIV-positive MSM may explain the increased syphilis incidence observed in San Francisco between 1998 and 2004. Our model results may have important implications for MSM not only in the US. Public health recommendations on HIV serosorting as an HIV harm reduction strategy should take into account the potential unintended consequence of increasing the prevalence of other STIs.