Polygyny and symmetric concurrency: comparing long-duration sexually transmitted infection prevalence using simulated sexual networks.

OBJECTIVE: To compare the effects of polygyny (only men can form concurrent partnerships) and gender-symmetric concurrency (both genders can form concurrent partnerships) on prevalence of long-duration sexually transmitted infections (STIs) using a dynamic stochastic network model. METHODS: We modelled two pairs of scenarios: polygyny and gender symmetry at higher and lower levels of network concurrency (measured by the average number of concurrent partnerships per partnership). The same level of sexual activity was modelled in all scenarios (measured by mean per capita partnership incidence and per capita number of sex acts). Partnership duration and network concurrency were constant within each of the polygyny/symmetry pairs. Infections that mimicked characteristics of herpes simplex virus type 2 (HSV2) and HIV were introduced onto the networks separately. The mean prevalence 100 years after introduction for the HSV2-like infection and 30 years after introduction for the HIV-like infection were calculated over 1000 model iterations. RESULTS: Prevalence of both simulated STIs was significantly lower in the polygyny scenarios than in the symmetry scenarios. At lower concurrency, polygyny resulted in a relative reduction in HSV2-like infection prevalence of 19% (95% CI 15 to 23) compared to gender symmetry. At higher concurrency polygyny led to a relative reduction of 20% (16 to 23). The relative reduction in prevalence of the HIV-like infection after 30 years was 14% (10 to 17) at lower concurrency and 8% (5 to 11) at higher concurrency. CONCLUSIONS: Polygyny can result in lower STI prevalence compared to populations where both genders practise concurrency. Further work is required to explore whether this reduction is observed when modelling more realistic populations and infection characteristics.