The Zika circuit

The Zika circuit

For much of its known existence, the mosquito-borne Zika virus was limited to parts of Africa and Asia. It took 50 or more years to reach the Americas (first detected in 2015), but then its pace quickly accelerated, spreading throughout the Western Hemisphere within about a year. Clearly, our vast travel networks and heavily urbanized centers bring together the essential components for transmission – the virus, the hosts (human), and the vectors (the human loving mosquitoes, Aedes aegypti). But yet, some fundamental questions remain. Why was Zika virus transmission quite intense on some Caribbean Islands and why was most of the southern US spared from local outbreaks?

Figure 1: Air travel routes used to construct the network in the ‘Zika circuit’. The virus can move between country nodes along the network routes. The amount of virus moving along each route depends on the travel volume and outbreak size at origin. Whether or not the virus becomes established at each node depends on local factors – like mosquitoes, humans, and economics.

To investigate the drivers of Zika virus spread, my friend and colleague, Lauren Gardner, took more of a mechanical approach than a biological one. Her solution was to view the Americas as a network, similar to how an electrical engineer might view a circuit board. Each country represents a node connected by a circuit of air travel routes (figure 1). The idea is that the virus can move along the routes to different nodes and the amount of movement is determined by 1) the air travel volume on that particular route and 2) the amount of transmission at the origin. Whether or not the virus becomes successfully established at the recipient country node depends on local conditions, such as mosquito and host density and its economic status. With this information, we asked: What factors best support the structure of  the network, given the local outbreaks that occurred at each node?      

To answer this question, we needed an array of data about the epidemic. This is where Mortiz Kraemer, Karthik Gangavarapu, and myself came in, to provide data about Zika outbreak sizes, the timing of spread, and vector abundance. Also, as part vector biologists/ epidemiologists/ computational biologists (see artist’s rendering), we provided alternative interpretations of what went into and out of the network model. Although we recognize many limitations with our approach, as with all modeling studies, we are quite satisfied with the end results, many of which contradict popular misconceptions.

At the top of this list, our results challenge the idea that infected travelers are the primary instigators of local Zika virus transmission. The debate over this idea hit its peak during the 2016 Rio Olympics in Brazil, where some wanted to ban the Games in fear of increased travel potentially sparking new outbreaks across the globe. In our ‘Zika circuit’, however, we found that air travel volume was the least impactful factor (although it is obviously still required). Rather, our results suggest that local factors are more to blame. As a prime example, the US received many times more Zika infected travelers than any of the Caribbean nations, yet the large outbreaks on these islands dwarfed the relatively small and contained Zika outbreaks in the US (Texas and Florida). Moreover, the local factor best associated with Zika virus transmission was its economic status – namely, its gross domestic product [GDP] per capita.

Figure 2. An urban slum in Brazil. The primary vector of Zika virus, Aedes aegypti, love to breed in small containers filled with rain water that litter these areas. Poor housing conditions do not effectively seperate the indoor from the outdoor environments. The combination makes for a large mosquito population with easy access to hosts – exposing the humans to more viruses. Photo credit: Albert Ko

The GDP is not directly involved in transmission, but it can be used as a measure of how often mosquitoes and humans come into contact. People of lower income status without the protection of government support systems can live in some pretty desperate conditions. In this scenario, where  urban slum rubbish provide excellent mosquito breeding habitats and homes often do not have air conditioning and/or window screens to keep the mosquitoes out, people are more likely to get exposed to Zika virus (figure 2). This hypothesis is supported by seroprevalence data (looking for antibodies from people previously infected) showing a higher Zika burden in lower socioeconomic areas of Brazil. Thus, as with many tropical diseases, Zika disproportionately affects the most economically disadvantaged and it should be considered a disease of poverty.

Ultimately, this was a fun collaboration that hopefully revealed some helpful insights. From a personal standpoint, this helped me view epidemics from a different vantage and will open up new opportunities for future research.

The paper:

Gardner LM, Bóta A, Gangavarapu K, Kraemer MUG, Grubaugh ND
Inferring the risk factors behind the geographical spread and transmission of Zika in the Americas
PLOS Neglected Tropical Diseases 12(1), e0006194 (2018)       

Interested in studying the emergence of mosquito-borne viruses, like Zika? See our job post.


Leave a Reply

Your email address will not be published.