Climate Change Is Driving Dengue Outbreaks. A New Study Shows Just How Much

Dengue has long been understood as a disease of the tropics, flaring where heat and humidity give mosquitoes an advantage. Yet transmission also has a more subtle signature, rising most sharply not in the hottest places on Earth, but where warming nudges temperatures into a deceptively mild range between 20 and 28°C. Those conditions now occur more often in many parts of the world, and a new study shows how strongly the dengue virus responds.

Researchers led by Marissa Childs and Erin Mordecai at Stanford used a set of causal inference tools more common in economics than in epidemiology to quantify how climate change is affecting dengue incidence. Their conclusion is difficult to ignore. Across 21 countries in the Americas and Asia, roughly 18 percent of dengue cases today are attributable to anthropogenic climate warming. In public health terms this is not a small effect. It means that climate conditions already experienced by hundreds of millions of people have shifted enough to intensify the burden of one of the world’s fastest growing infectious diseases.

I spoke with Childs and Mordecai soon after the study appeared in the Proceedings of the National Academy of Sciences, one of the most prestigious scientific journals in the world. Like many readers, I was impressed by the scale of the work. The team assembled more than five million monthly records of dengue incidence across 6639 locations, each with its own reporting system and idiosyncrasies. Childs told me this effort alone consumed years. “Everyone’s storing their data in a different way. Getting all of those data sets together and wrangled in a usable format was a big, big task,” she said with a quiet pride familiar to anyone who has spent long hours cleaning real-world data.

To isolate that signal, the team applied a panel regression framework that controls for nearly everything that varies across places and seasons except for the short-term temperature anomalies they cared about. This approach strips away confounding information, controlling for long-term changes like urban growth or shifting virus types. It removes seasonal cycles within countries and the fixed characteristics of each location. “What we’re left with is actually a very small portion of the temperature signal, but it’s the portion of the temperature signal that we actually think can be traced, that we can link to dengue.” Linking those anomalies reveals the causal effect of temperature.

The resulting temperature–incidence relationship is nonlinear. Dengue incidence rises steeply as temperatures warm from roughly 15°C into the mid-20s, peaking at 27.8°C, then declining at higher temperatures. For most places in the study the greatest risk comes not from extreme heat but from warming into the optimal transmission zone. This explains why some of the largest projected increases occur in relatively cool regions of South America. Meanwhile, very hot places like Cambodia see little effect from further warming because they already sit above the temperature that favors transmission.

As a researcher who works on the dynamics of infectious diseases, I was struck by the care the authors took to show that their estimates reflect temperature rather than confounding factors. Dengue is shaped by many hard-to-observe processes, including population immunity and interactions among different viral serotypes. The team used several strategies to test the robustness of their results. In simulated dengue systems with known temperature sensitivities, the econometric model recovered the correct thermal optimum. In the observational data, they controlled for recent dengue history, proxies for immunity and socioeconomic differences. In every case the temperature signal persisted.

“I think there’s a bunch of reasons to think it’s a lower bound,” Childs told me. Among these, the study focuses only on places where dengue is already endemic. It does not capture the expansion of dengue into new regions. Nor does it include many large countries where dengue occurs, most notably India and all of Africa. The authors chose not to extrapolate beyond the 21 nations with high quality data. This makes the work conservative by design.

Still, the implications are sobering. Under a moderate emissions pathway the authors project a 49 percent increase in dengue incidence by midcentury. Under a high emissions scenario the increase reaches 76 percent. About 262 million people already live in places where dengue is expected to more than double. Because reported cases capture only a fraction of infections, the attributable burden would be far larger if underreporting were to be considered.

Dengue differs from many other climate-related health risks in the temperature range at which warming begins to matter. In conversation, Mordecai emphasized this point. “Those are temperatures that are already really common,” she said. “They’re being experienced right now. And that’s why we’re picking up on such strong impacts of climate change already, because you don’t have to wait until you get to these extremes.”

There is also a deeper scientific story here that resonated with me. Ten years ago Mordecai and colleagues published mechanistic models predicting how dengue transmission should depend on temperature based on mosquito physiology. Those predictions match almost exactly what Childs and colleagues have now inferred from real-world data spanning two continents. It is rare to see theoretical biology borne out so neatly in a complex global system.

For policymakers the message is clear. Dengue is already responding to climate change in measurable ways. Adaptation will need to keep pace with rising temperatures long before extreme heat becomes the dominant concern. And mitigation matters. The difference between low and high emissions scenarios represents millions of infections each year.

This study opens the door to similar attribution work for other vector-borne diseases. For now it provides one of the clearest demonstrations that climate change is not only a future threat to human health, but one that operates through everyday conditions rather than extreme events, and is actively reshaping the infectious disease landscape.