SYMP 8-5
Can large scale human processes explain fine-scale heterogeneities in endemic dengue transmission?
Heterogeneity in contact between hosts increases the explosiveness of pathogen transmission, but its role in endemic diseases, when some hosts are immune, is less well understood. This question is crucial for assessing how ecological processes impact mosquito-borne viruses, as heterogeneity in infectious contact largely derives from the biotic and abiotic interactions of mosquitoes. Here we combine a field study of the prevalence of dengue viremia in 10, 2-block patches located in highly endemic neighborhoods of Colombia, with an agent based model, in order to assess the causes and consequences of endemic heterogeneity in dengue transmission.
Results/Conclusions
In an unprecedented finding, we confirm the potentially explosive nature of endemic dengue transmission. While 9 study patches had relatively low levels of dengue viremia, consistent with previous studies, in one patch, located in the city of Armenia, we demonstrate simultaneous DEN2 presence in 22 of 23 children living within a 40m radius. Our model, parameterized with multi-year data on Aedes aegypti recruitment from the same neighborhood, demonstrates that while increased host- vector contact favors such bursts during the initial invasion of a dengue serotype, larger scale human dynamics drive their occurrence 10-20 years post-invasion. During this period transmission bursts will play a critical role in dengue persistence in areas that maintain high densities of susceptible humans and will occur most frequently at intermediate levels of social connectivity with external areas. These results are consistent with our observations that the high viremia patch was located in one of the more peripheral neighborhoods of Armenia that receives migrants from non-endemic rural areas, and had the youngest age structure and lowest mean residence time in the neighborhood of all 10 patches. We show that vector control can significantly reduce bursts of dengue transmission in areas with high rural-to-urban migration, high fertility and low viral introduction, but has little impact and can even be counterproductive in areas where frequent human introduction of virus maintains high levels of human immunity. These data suggest that the successes and failures of dengue management in modern hyper-endemic settings depend critically on geographic and temporal variation in human life histories and social networks.