Environmental conditions and animal hosts shape the spatial transmission risks of echinococcosis in China.

Echinococcosis poses a major public-health challenge and substantial socioeconomic burdens in pastoral regions. Yet the spatial transmission risks in China and their drivers remain insufficiently understood. Using a national cross-sectional survey of human and animal echinococcosis (2012-2016), we applied ensemble machine-learning models to identify environmental determinants of disease distribution, predict infection risk in animal hosts, and project human transmission risk. Model-predicted animal-host infection risk increased under humid, low-temperature conditions. Human cystic echinococcosis (CE) shows positive, non-linear associations with predicted infection risk in dogs and cattle, whereas human alveolar echinococcosis (AE) is positively, non-linearly associated with predicted infection risk in rodents and dogs. High-risk areas for sheep and dog infection are projected mainly in Qinghai, northwestern Tibet, Xinjiang, northern Inner Mongolia, and central Gansu, with broader extent than for cattle or rodents. The Tibetan Plateau emerges as a hotspot for human CE, while human AE risk concentrates along the Qinghai-Sichuan-Tibet border regions. These findings clarify environment-host-human linkages and provide a framework for targeted surveillance and control of echinococcosis under climate change.