The American Academy of Microbiology, an elected body of the field’s most distinguished scientists, has released a report with a pointed question at its center: we know climate change is spreading infectious disease, but how well have we actually measured the damage? The answer, according to a colloquium of leading epidemiologists, microbiologists and climate scientists, is not well enough. I was a participant in that colloquium, and even among specialists the sparseness of the evidence was sobering.

Warmer temperatures expand the range of mosquitoes that carry dengue. Shifting rainfall alters the ecology of cholera . The area suitable for West Nile virus in Europe is expanding . Everyone in the field agrees on the direction. Where the science gets harder is putting a number on it.

Counting matters because without knowing how big the problem is, one cannot decide how much to spend on it, where to direct resources or whom to hold accountable. The difficulty is that you cannot count something that did not happen. We only get one version of history, one set of dengue cases, one trajectory of warming. There is no control Earth.

Enter attribution science. Attribution science does something that correlation studies cannot. It uses global climate models, the same coupled atmosphere-ocean simulations that underpin IPCC projections, to construct a counterfactual: a version of Earth’s climate history with natural variability intact but anthropogenic greenhouse gas emissions removed. Researchers then run disease transmission models under both the real climate and the counterfactual and compare the outcomes. The method borrows from climate science, where attribution has been used for over two decades to determine how much of a heat wave or hurricane is explained by human activity. Applied to infectious disease, it answers a specific question: of the dengue cases in Southeast Asia or the lengthening West Nile virus season in New York, what fraction would have happened anyway?

The report is blunt about this: “climate is not merely a background condition but a core driver of infectious disease dynamics.” Correlation can show that dengue cases and warming temperatures move together, but it cannot separate the climate signal from everything else that changes at the same time, such as urbanization, mosquito control spending or population growth. Attribution isolates the climate effect by holding those other factors constant in the model. The result is a number, say 18% of dengue cases, that policymakers can use to direct budgets, guide legal arguments and put public health resources where they will do the most good.

The primary attribution studies to date are all in mosquito-borne disease systems, where the link between temperature and transmission is best understood. A landmark 2025 study found that roughly 18% of the global dengue burden across 21 countries in the Americas and Asia was attributable to climate change, with projections of 49 to 76% increases by mid-century. (I have written about this in greater detail see here .) Another traced 60% of dengue cases in northwestern Peru after Cyclone Yaku in 2023 to climate-driven extreme rainfall. In New York State, a preprint from Stanford University that has not yet been peer-reviewed found that the West Nile virus transmission season has lengthened by 20 days over 25 years, a trend 6.4 times more likely under anthropogenic warming than without it.

For malaria , the picture is more geographically complex: climate warming has increased childhood malaria prevalence in the cooler highlands of southern and east Africa, where rising temperatures now permit transmission that was once impossible, but has decreased it in west Africa, where temperatures already exceed the biological optimum for the parasite.

Quantifying the actual contribution of climate change requires long-term data, mechanistic experiments, climate models and statistical frameworks that most disease systems lack. The report identifies fungal infections, waterborne pathogens and tick-borne diseases as systems where the groundwork has barely begun. Fungal pathogens are a particular gap. Candida auris , Valley fever and other climate-sensitive fungal infections are expanding rapidly, yet the mechanistic data needed for attribution studies barely exist. If attribution science is going to inform policy, it will need to reach beyond the diseases where the data are richest.

Public health systems worldwide are under strain from reduced funding, eroded trust and the lingering aftermath of the pandemic. Climate-sensitive diseases are already arriving in new places. Los Angeles County recorded its first sustained local dengue transmission in the fall of 2024. As the report warns, “disease burdens often change faster than studies can be conducted.”

The diseases are not waiting for the science to catch up.