Select Page

A Cooldown for Malaria Predictions

By Rob Mitchum // October 25, 2012

Malaria is often described as a tropical disease, mostly afflicting populations in sub-Saharan Africa and equatorial regions of Asia and the Americas. Due to climate change, those warm parts of the world are getting warmer, leading some experts to speculate that malaria will become an even bigger problem than the enormous current tally of approximately 200 million infections and 650,000 deaths a year. But a new model, published this month in Ecology Letters, that better reflects the complex influence of temperature on mosquito biology produced a surprising result: that malaria infection peaks at relatively mild temperatures several degrees below previous estimates. That could be good news for those in tropical regions…but maybe not such good news for the rest of us.

Models of malaria infection are based on parameters describing the biology and behavior of humans, mosquitoes, and the malaria parasite itself. Many of biological rates related to processes such as reproduction or maturation are temperature-dependent — especially so for a cold-blooded species like the mosquito. But rates don’t increase forever as the temperature rises; eventually, they reach a maximum limit and decline in what is known as a unimodal function (basically an upside-down “U”). Yet most models of malaria infection have been built upon the assumption that these pieces of mosquito biology are linearly related to temperature.

“In a lot of these previous models, at least one component was assumed to increase with temperature without any limits,” said Leah Johnson, a research professional at the University of Chicago. “For instance, reproduction of mosquitos would just keep on increasing as temperature increases, but physiological principles dictate that is not possible.”

Johnson is part of a team of collaborators from UChicago, UCSB, Penn State, UCLA, SUNY-ESF, and the US Geological Survey that decided to build a new model that more accurately reflects the complex relationship between mosquito biology and temperature. First, they gathered data from previous experiments that had studied mosquitos kept in laboratories at constant temperature environments, which controlled for fluctuations and non-temperature factors such as humidity. Combining these data, the team found that every mosquito-parasite trait they examined — from survival and bite rates to the development of both the mosquito host and the malaria parasite — followed a unimodal curve as temperature increased.

ele12015-fig-0002.pngThe team then plugged that more nuanced mathematical understanding of mosquito biology into a new-and-improved malaria model, replacing the linear relationships with unimodal ones. The resulting model predicted that the optimal temperature for malaria transmission was only 25 ºC (77 ºF) — five to seven degrees celsius below what the previous models predicted for peak malaria transmission. Above 25 ºC, in the ranges found in most tropical areas, the malaria infection rate was predicted to steeply decline.

“We are really excited about the new findings, because they tell us temperatures as low as 82°F may begin to slow malaria transmission,” says lead author Erin Mordecai, of the University of California Santa Barbara (UCSB). “Our study challenges the common assumption that hot temperatures simply speed up transmission.”

Faced with such an unexpected result, the collaboration tested their model with field observations on malaria transmission from Africa, which confirmed the cooler peak. The conclusion also agreed with research from the 1930’s that found that malaria infection peaks in the cooler months of autumn rather than summer. But if peak transmission is expected to be found at temperatures more akin to California than Central Africa, why don’t we see more infections in milder regions? The authors speculated that the socioeconomic advantages of milder and wealthier regions, such as screened windows, air conditioning, and more sophisticated insect control, are sufficient to mitigate the elevated threat.

But in the context of climate change, the tables could turn. The new model’s cooler peak could mean that tropical areas heavily afflicted by the disease will see declining malaria rates in the future as the mercury rises. But higher temperatures elsewhere in the world could push more moderate regions into the malaria danger zone, requiring more intensive efforts to eradicate the disease and the insects that carry it.

“One of the things that’s somewhat of a relief about this is that it predicts that the places that already have a lot of malaria maybe aren’t going to get worse as temperatures increase,” Johnson said. “A lot of places, in Africa for instance, are already at or above the peak. On the other hand, we may see that risk will increase more quickly in temperate areas than we expected. As temperature shifts, it could be that these areas need to start using a lot more resources to maintain control.”

More papers are due to be submitted soon by the collaboration as they refine their malaria model with additional mathematical methods and research on the relationship between mosquito biology and other environmental factors. Eventually, they hope that the model will help policymakers determine more effective strategies for distributing anti-malaria resources, using the predictions to focus prevention efforts on regions or times of the year where malaria transmission is expected to be at its highest. The model may also prove useful for looking at diseases other than malaria, said co-author Samraat Pawar, a post-doctoral researcher in the laboratory of Computation Institute faculty member Stefano Allesina.

“Our model structure is general, and can be applied to any cold-blooded species of disease vectors and parasites,” said Pawar, who was previously at the David Geffen School of Medicine in UCLA.So it will be useful for other researchers trying to predict effects of climate on other vector-borne diseases of humans and other animals.”