The Impact of Extreme Weather on Dengue Fever in Guangzhou, China: A Zero‐Inflated Negative Binomial Spatial Lag Analysis

Climate change intensifies extreme weather, which in turn influences infectious disease transmission. As a dengue fever (DF) hotspot, Guangzhou lacks research on how extreme weather characteristics and spatial factors interact to shape DF patterns. This study analyzed DF distribution in Guangzhou from 2017 to 2019, using a zero-inflated negative binomial spatial lag (ZINB-SAR) regression model to assess the effects of daytime heatwaves (DHW), nighttime heatwaves (NHW) and extreme precipitation (EP) on DF. Results revealed that DF cases were predominantly clustered in central urban areas, with an epidemic season from May to November. The ZINB-SAR model outperformed negative binomial regression and spatial econometric models, with all spatial effect coefficients significantly positive. Analysis of lagged effects showed that each additional DHW event increased DF cases by up to 10.80% (95% CI: 6.22%-15.59%) at a 2-month lag, while NHW events increased DF by 2.73% (95% CI: -1.59%-7.23%). Threshold analysis indicated DHW intensity shifted from promoting to inhibiting DF between 0.66°C and 0.76°C, while NHW intensity transitioned between 0.95°C and 2.28°C. EP demonstrated the strongest effects at a 3-month lag, increasing DF cases by 12.05% (95% CI: 9.03%-15.17%), although its intensity was not statistically significant. Seasonal and spatial variations in DF incidence were evident. In conclusion, DHW and EP impacts were primarily driven by event frequency rather than intensity, whereas NHW effects were more dependent on intensity. These findings highlight the complex spatiotemporal interplay between extreme weather and DF in Guangzhou, providing critical evidence for developing targeted climate-adaptive disease control strategies.