Every year, a handful of Americans develop a mysterious illness that begins with fatigue, fever, and muscle aches before rapidly filling their lungs with fluid. About one in three will die. The culprit is hantavirus pulmonary syndrome, a rare but devastating disease spread not by other sick people, not by contaminated water, but by breathing in microscopic particles of infected rodent droppings.

Hantavirus has lurked in American landscapes since at least 1993, when an outbreak in the Four Corners region of the American Southwest first revealed its presence in the continent. Since then, 864 human cases have been recorded across the United States. That number sounds small until you consider the mortality rate — roughly 35 percent — which makes hantavirus one of the deadliest infectious diseases circulating on U.S. soil today.

Despite decades of surveillance, no one had produced a comprehensive national risk map drawing on climate, land use, rodent diversity, and socioeconomic vulnerability together. A new study published in Transboundary and Emerging Diseases does exactly that, and its findings carry clear implications for where public health resources should be directed — and where the threat may be growing.

Mapping Where Risk Lives

To understand which places in the contiguous United States carry the highest long-term risk of hantavirus exposure, researchers from Los Alamos National Laboratory and the U.S. Centers for Disease Control and Prevention applied a modeling technique called ecological niche modeling to 30 years of human case data.

Ecological niche modeling is a statistical approach originally developed to map the geographic ranges of plant and animal species. Here it was repurposed to map the range of hantavirus risk itself — treating each recorded human case as a signal that local conditions were suitable for transmission.

The dataset used in the analysis covered 821 cases reported from 1993 through 2022. After removing cases linked to exposures outside the country, to a separate strain called Seoul virus associated with pet rats, and to cases lacking location information, the team worked with 431 cases precisely filtered to minimize the distortion caused by outbreak clustering.

Those cases were fed into the model alongside 29 environmental variables spanning four categories: climate (seven variables including precipitation, temperature, and soil moisture), land cover (16 land use types from shrubland to developed suburban spaces), rodent species richness, and four measures of social vulnerability drawn from the CDC's Social Vulnerability Index.

The result was a set of risk maps at four kilometer resolution, covering the entire contiguous United States.

The West Is Different

The most striking finding is also the most consistent: the western United States carries dramatically higher long-term hantavirus risk than the east. Of the 431 cases analyzed, 400 — fully 93 percent — occurred in the western half of the country.

The model explains why. Three variables together accounted for nearly half of all predictive power in the national model: land classified as "developed, open space," land classified as "developed, low intensity," and mean temperature. In the western U.S. model, those same three variables explained an even larger share — 63 percent — of the geographic pattern of risk.

Areas of highest predicted risk included central Colorado, northern New Mexico, Utah, southern California, eastern Washington, and the Snake River Valley in Idaho. Risk was also elevated along highway and interstate corridors and in suburban fringe areas throughout the west.

Temperature mattered in a specific direction: lower mean temperatures were associated with higher risk. Precipitation showed an inverse relationship too — the drier the location, the higher the predicted risk. When the researchers ran a climate-only model, precipitation alone accounted for 26 percent of predictive importance, followed closely by temperature range and maximum temperature.

The explanation likely involves the virus's survival outside a host. Hantaviruses can persist in rodent excrement for up to 15 days, and dry conditions allow infected material to accumulate and become airborne more easily. In the wetter eastern United States, rainfall may wash contaminated droppings into drains and ditches before humans ever come into contact with them. In the arid west, infected material can linger, desiccate, and become dust — inhaled during cleaning, construction, camping, or simply moving through the wrong space at the wrong time.

The Fringe Effect

One of the more nuanced findings involves what researchers call fringe habitats — transitional zones where one type of land cover gives way to another. Human cases of hantavirus appeared disproportionately in these edge environments, near but not within shrubland, evergreen forest, grassland, and cropland.

This pattern appears across multiple models. In the eastern United States, areas with smaller proportions of pasture or woody wetland coverage carried higher risk than areas dominated by those land types. In the west, similar transitions around forests, shrubs, and crops showed the same signal.

The finding echoes what ecologists have observed for other emerging infectious diseases: the boundary between wild habitat and human activity is where spillover events concentrate. As development pushes into previously undisturbed land at low intensity — suburban sprawl, rural outbuildings, weekend cabins — people enter the ecological territory of the deer mouse, the primary carrier of Sin Nombre virus, the strain responsible for the vast majority of U.S. cases.

Deer mice are generalists. They thrive across an enormous range of habitats in the west, and unlike some rodent species, they do venture into buildings — particularly structures that border woodland or open land. They are less common in dense city centers. The risk, in other words, is concentrated at the margins: where the city ends and the countryside begins.

Rodents, Richness, and an Unexpected Signal

The role of rodent species diversity in hantavirus transmission is scientifically contested. Some studies suggest that higher diversity dilutes the pathogen — when a deer mouse shares habitat with many other species, the virus has fewer opportunities to spread among its primary host. Other evidence points toward an amplification effect, where diversity increases transmission.

The new analysis found a positive association between rodent species richness and hantavirus risk, meaning areas with more rodent species were generally at higher predicted risk. This held across the national model, the western U.S. model, and the land use focused model.

The researchers caution that this result may partly reflect a geographic coincidence: the Four Corners states — Arizona, Colorado, New Mexico, and Utah — account for a large share of U.S. hantavirus cases and also happen to have some of the country's highest rodent diversity. Climate and rodent richness are correlated in ways that make it difficult to fully separate their contributions.

Still, the finding gains support from recent field research. A 2024 study in eastern New Mexico identified Sin Nombre virus in 16 different rodent species using sensitive genetic testing. A 2025 analysis documented hantavirus antibodies across 15 rodent species, including six previously unrecognized hosts. The picture that emerges is of a virus with a broader ecological footprint than traditionally assumed — not confined to the deer mouse alone, but potentially circulating through entire rodent communities.

Who Is Most Vulnerable

The social dimension of hantavirus risk is less studied but increasingly important. When researchers built a model using only the four social vulnerability variables, the most important single predictor was household composition — a measure that captures the proportion of households with dependent children under 18, elderly residents, single parents, and people with disabilities.

This was unexpected. Children under 18 account for only about 10 percent of U.S. hantavirus cases, and there is no strong existing evidence for age-stratified risk in the general population. Why household composition would rank as the dominant social predictor is not yet clear, and the researchers flag it as a priority for further investigation.

What is clearer is that socioeconomic vulnerability tracks with hantavirus risk in the west. The model found a nonlinear positive relationship between social vulnerability indicators and predicted risk — meaning that as communities become more economically and socially marginal, risk tends to rise. Poor housing conditions, homes with entry points for rodents, and limited access to healthcare all likely contribute, though disentangling these effects requires more targeted research.

Interestingly, no social vulnerability variables were significant predictors in the eastern U.S. model, suggesting that in the east, the disease ecology itself — different rodent species, different hantavirus strains, different landscape types — overrides the social gradient that shapes western risk.

What Climate Change May Bring

The climate signal in the data carries a forward-looking warning. The southwestern United States is already the driest part of the country, and current climate projections suggest it will become drier still. Meanwhile, temperatures are expected to increase across the west, and dust storms — which can aerosolize contaminated material — are projected to intensify.

The model's findings imply that these trends could push hantavirus risk higher in the southwest over coming decades, potentially expanding the geographic footprint of transmission into areas not historically affected. Canada has already recorded 143 cases of hantavirus cardiopulmonary syndrome as of early 2020, most concentrated in the western part of the country, and warming conditions could extend suitable habitat further north.

The eastern United States, by contrast, is projected to become wetter — which the model suggests would maintain the natural suppression of hantavirus transmission in that region.

Why This Matters for Public Health

For a disease that kills roughly one in three people it infects, hantavirus receives relatively little public attention. Part of the reason is its rarity — fewer than 900 confirmed cases across three decades of national surveillance. But rarity does not mean randomness. The new risk maps reveal that exposure risk is geographically structured, ecologically predictable, and tied to identifiable social and environmental conditions.

The maps have practical applications. Public health officials in high-risk regions can use them to target awareness campaigns, rodent control programs, and housing inspections toward the communities and landscapes most at risk. Physicians in the western United States can apply them to inform differential diagnosis decisions. Emergency planners can use them to anticipate where hantavirus cases are most likely to emerge or be underreported.

The researchers also highlight that some regions with high modeled risk have historically low case counts — not necessarily because exposure is rare there, but potentially because surveillance is inadequate. Hantavirus is difficult to diagnose without a high degree of clinical suspicion, and milder or atypical cases may be missed.

The disease does not require contact with a visibly sick animal. A dusty shed. A cabin opened for the season. A hiking trail through dry scrubland. In the right landscape, under the right conditions, exposure can be nearly invisible — which is precisely why knowing where those conditions cluster matters so much.

Credit & Disclaimer: This article is a popular science summary written to make peer-reviewed research accessible to a broad audience. All scientific facts, findings, and conclusions presented here are drawn directly and accurately from the original research paper. Readers are strongly encouraged to consult the full research article for complete data, methodologies, and scientific detail. The article can be accessed through https://doi.org/10.1155/tbed/7126411