Cardiology Research and Practice · Nov 10, 2026

Short‐Term Interactive Effects of Temperature and Air Pollution on Cardiovascular Visits: A Machine‐Learning Study in Hohhot, China

By Mao X, Wang T, Jiao X et al.

Introduction: The synergistic effects of ambient temperature and air pollutants pose a substantial public health burden related to cardiovascular disease (CVD). However, their high dimensionality, nonlinearity, and temporal lag structures make them difficult to characterize with conventional statistical methods. Methods: We developed a machine‐learning framework to evaluate the short‐term interactive effects of ambient temperature and six major air pollutants (PM ₂ . ₅ , PM ₁₀ , SO ₂ , NO ₂ , CO, and O ₃ ) on CVD visits. Using 508,077 records from tertiary hospitals in Hohhot, China (2016–2024), we matched daily CVD visits to these environmental data. A high‐order polynomial ridge regression modeled nonlinear exposure interactions, quasi‐Poisson regression captured lag structures, and a random forest assessed overall model performance. Results: High temperatures significantly amplified the cardiovascular risks of PM ₂ . ₅ and SO ₂ exposure, with PM ₂ . ₅ showing a cumulative percentage change of 30.96% at lag07 (95% CI: 30.86%–31.05%, p < 0.001). Conversely, the adverse effects of PM ₁₀ , NO ₂ , CO, and O ₃ were more pronounced under low temperatures. The random forest model achieved high predictive accuracy (cross‐validated R ² = 0.741). Conclusion: Air pollution and temperature jointly influence CVD health risks. Machine learning improves characterization of these complex relationships. These findings support the development of early warning systems and targeted interventions for CVD prevention under varying temperature conditions.

doi.org/10.1155/crp/9938043Open accessPubMed 43017284Populations

AgingResearch.News