But the story is different at higher elevations, including the western Kenya highlands, the Usambara Mountains of Tanzania, the highlands of Ethiopia, the Ruwenzori Mountains of Uganda, and the mountains of Indonesia and New Guinea. The good news in such areas is that most residents have not suffered from malaria; the bad news is the same. Because their bodies have never been exposed to the parasite, their immune systems respond too slowly to stop the infection during its early stages. And they, unlike lowland residents, usually lack two genetic traits that make people less susceptible to malaria. This genetic and immunological vulnerability mean that when malaria does hit highland residents, it hits harder, creating waves of disease that lay waste to vulnerable Africans, just as smallpox laid waste to vulnerable Native Americans during early colonial times. People in highland areas account for 12.4 million cases each year—just 2.5 percent of the global total—but the 150,000 annual malaria deaths in highland areas are 12 – 25 percent of the annual worldwide total. One in five residents of East Africa, or about 125 million people, live in highland areas susceptible to malaria epidemics.
* * *
To prevent an epidemic, it pays to predict it, and to do that one first has to understand how epidemics tend to unfold. Typically, cases trickle in at first, then skyrocket, then peak and fall off. Plotted over time, the result is a bellshaped curve—the same curve that's used to plot the distribution of student grades in a class. Health officials typically detect an epidemic only when the first rash of cases has been diagnosed and numbers are already climbing. By the time they can institute protective measures, many of the people who will be sickened are already infected, and it's too late to stop the spread. During malaria epidemics in the East African highlands, the disease can infect up to 60 percent of a village population, overwhelming local health workers. It's even worse when the parasite is drugresistant: thousands get sick; hundreds can die.
Armed with his laptop computer, in the late 1990s Githeko began modeling the effect of climate and weather on malaria transmission. He wanted to use local weather data to predict epidemics early enough to ship medicine and health workers to villages that would be affected, to distribute bed nets, and to spray insecticides that kill malarial mosquitoes. But predicting epidemics of infectious disease is no routine task.
Githeko knew that warmer temperatures—within a range—made it more likely that mosquitoes would survive and bite more often, and that it they allowed parasites to multiply and mature more quickly in mosquitoes. He also knew that rainfall created the puddles and standing water that are ideal breeding spots for mosquitoes and could quickly make them more abundant. By predicting temperatures and rainfall over a period of months, he could, in theory, project where and when an epidemic was most likely to arise.
Without any of the hightech capabilities grant money would have provided, Githeko was able to use a simple spreadsheet program—and historical temperature and rainfall data for the western Kenya highlands, as well as data from published studies of mosquitoes, malaria parasites, and past epidemics—to develop a mathematical model that correctly predicted past outbreaks in the region, thereby verifying the accuracy of his model.
The model also made a startling new prediction. By the late 1990s, when Githeko did his analysis, the IPCC already had reported that the Earth had warmed 0.6 degrees C (1.1 degrees F) through the 20th century and had predicted that temperatures would continue to rise by at least several degrees more through the 21st. Githeko's model indicated that a temperature change of just 2 degrees C could trigger malaria epidemics in many oncesafe highland regions of East Africa. This meant that global warming would eventually make millions of people in malariafree parts of East Africa—including his family in the highlands of central Kenya—vulnerable to the disease. Though alarming, the threat seemed distant; climate scientists thought it would take decades for the region to warm that much.