Extreme Urban Temperature Exposure and Preterm Birth: Spatial-Temporal Risk Zone Prediction Using Machine Learning Models

Yuxiao Wang; Lingchen Bu; Zanmei Wei; Yang Cheng; Ling Feng; Shaoshuai Wang

doi:10.1016/j.envres.2025.122230

Extreme Urban Temperature Exposure and Preterm Birth: Spatial-Temporal Risk Zone Prediction Using Machine Learning Models

Environ Res. 2025 Jun 25:122230. doi: 10.1016/j.envres.2025.122230. Online ahead of print.

Authors

Yuxiao Wang¹, Lingchen Bu¹, Zanmei Wei¹, Yang Cheng², Ling Feng³, Shaoshuai Wang³

Affiliations

¹ Faculty of Geographical Science, Beijing Normal University, 100875, Beijing, China.
² Faculty of Geographical Science, Beijing Normal University, 100875, Beijing, China. Electronic address: chengyang@bnu.edu.cn.
³ Department of Obstetrics and Gynecology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan 430030, China.

PMID: 40578740
DOI: 10.1016/j.envres.2025.122230

Abstract

This study investigates temperature impacts on preterm birth (PTB) using residential address GPS coordinates for 311,972 pregnant women in Wuhan, China, coupled with daily environmental data. We developed a machine learning model to analyze the impact of environmental exposure on PTB. Results show PTB risks increase with temperatures below 14°C or above 21°C, excessive temperature variability, and acute exposure to extreme weather. Spatial analysis revealed heat-related risk zones concentrated in urban heat island areas, while cold-related risks were more widespread. This research provides novel insights into spatiotemporal patterns of temperature-related PTB risk, offering evidence-based recommendations for urban health planning and climate adaptation strategies. The integration of machine learning and spatial analysis represents a significant advancement in environmental health research methodology.

Keywords: extreme temperatures; machine learning; preterm birth; temperature exposure risk zones; urban heat island effect.