A global analysis of 58,319 outbreak events across 32 emerging infectious diseases has found that the geography of reported outbreaks reflects a complex combination of ecological change, human activity and access to healthcare. One of the clearest patterns was not an environmental pressure itself: outbreak reporting fell sharply as travel time to a health facility increased.
A global test of outbreak drivers
Researchers brought together geolocated human outbreak and case records for 32 environmentally linked emerging infectious diseases. The dataset included zoonotic, vector-borne and environmentally mediated infections, ranging from dengue and Lyme disease to Ebola, mpox, plague and several mosquito-borne arboviruses.
In total, the analysis incorporated 58,319 outbreak event records. The researchers systematically tested 16 social and environmental drivers using a standardised geospatial modelling framework. These included forest and cropland cover, landscape fragmentation, biodiversity loss, mining, urbanisation, livestock density, long-term temperature and precipitation change, socioeconomic vulnerability and travel time to healthcare.
Because outbreak records are vulnerable to detection and reporting bias, the models were designed to account for spatial differences in where cases are likely to be identified and documented. The researchers generated population-weighted background locations and used geospatial logistic regression to compare conditions at outbreak locations with conditions experienced by populations across each disease’s study region.
Healthcare access produced one of the strongest common patterns
The most consistent finding concerned access to healthcare. Across the diseases examined, the odds of an outbreak being reported declined by a median of 32% for every additional hour of motorised travel time from the nearest healthcare facility. The size of this decline varied substantially between diseases, ranging from 1.2% to 96.7%.
This result does not mean that infections necessarily become less common in remote areas. Instead, it highlights an important surveillance problem. People living farther from clinics and hospitals may have fewer opportunities to receive a diagnosis, while remote health facilities may have less capacity to identify uncommon pathogens. The observed map of outbreaks can therefore differ from the underlying map of infection risk.
Mosaic landscapes repeatedly appeared as areas of elevated risk
Outbreak risks were often highest in mosaic landscapes where human populations and livestock live alongside forests and fragmented ecosystems. These mixed landscapes can create interfaces between people, domestic animals, wildlife and disease vectors.
Several vector-borne diseases shared common associations with these landscape conditions. Long-term declines in precipitation also showed important effects across diseases including dengue, Lyme disease and zoonotic arboviruses.
However, the researchers did not find a universal environmental explanation for emerging disease outbreaks. Directly transmitted zoonoses such as Ebola and mpox shared relatively few common drivers. The effects of deforestation, climate warming, agricultural intensification and other anthropogenic pressures differed substantially between disease systems.
Why the findings matter
The results challenge simple explanations in which one environmental pressure is treated as the dominant global cause of emerging infectious disease. Different pathogens depend on different hosts, vectors, transmission routes and ecological conditions, meaning interventions that work for one disease may not transfer directly to another.
The findings also reinforce the importance of health-system capacity as part of pandemic prevention. Stronger local diagnostic services and surveillance can improve the detection of outbreaks in remote populations and help distinguish genuine differences in disease occurrence from differences in reporting.
The researchers argue that ecosystem-based public-health interventions should therefore be based on evidence specific to each disease system and paired with investment in health systems and One Health pathogen surveillance.
Important limitations
The study is observational and geospatial, so associations between environmental conditions and outbreak locations should not automatically be interpreted as direct causal effects. The underlying outbreak datasets also vary in their geographic coverage, temporal coverage and diagnostic quality.
Detection and reporting bias remain substantial despite the researchers’ efforts to model them. Some pathogens of particularly high concern, including several bat-borne epidemic viruses, had relatively sparse data. The analysis also focused on factors associated with where outbreaks occur, rather than the separate processes that determine whether an outbreak expands into a large epidemic or pandemic.
Finally, broad global variables cannot capture every local behavioural, ecological or institutional factor affecting transmission. The substantial differences between diseases are therefore as important as the common patterns identified across them.
Source Information
Study: The anthropogenic fingerprint on emerging infectious diseases
Authors: Rory Gibb, Sadie J. Ryan, David M. Pigott and colleagues
Journal: Nature
Published: 23 September 2026
DOI: 10.1038/s41586-026-11058-6
Study type: Global geospatial analysis of outbreak records
Data: 58,319 outbreak event records across 32 emerging infectious diseases
