How do temperature and precipitation drive dengue transmission in nine cities, in Guangdong Province, China: a Bayesian spatio-temporal model analysis

Dengue remains an important public health issue in South China. In this study, we aim to quantify the effect of climatic factors on dengue in nine cities of the Pearl River Delta (PRD) in South China. Monthly dengue cases, climatic factors, socio-economic, geographical, and mosquito density data in nine cities of the PRD from 2008 to 2019 were collected. A generalized additive model (GAM) was applied to investigate the exposure–response relationship between climatic factors (temperature and precipitation) and dengue incidence in each city. A spatio-temporal conditional autoregressive model (ST-CAR) with a Bayesian framework was employed to estimate the effect of temperature and precipitation on dengue and to explore the temporal trend of the dengue risk by adjusting the socioeconomic and geographical factors. There was a positive non-linear association between the temperature and dengue incidence in the nine cities in south China, while the approximate linear negative relationship between precipitation and dengue incidence was found in most of the cities. The ST-CAR model analysis showed the risk of dengue transmission increased by 101.0% (RR: 2.010, 95% CI: 1.818 to 2.151) for 1 °C increase in monthly temperature at 2 months lag in the overall nine cities, while a 3.2% decrease (relative risk (RR): 0.968, 95% CI: 0.946 to 0.985) and a 2.1% decrease (RR: 0.979, 95% CI: 0.975 to 0.983) for 10 mm increase in monthly precipitation at present month and 3 months lag. The expected incidence of dengue has risen again since 2015, and the highest incidence was in Guangzhou City. Our study showed that climatic factors, including temperature and precipitation would drive the dengue transmission, and the dengue epidemic risk has been increasing. The findings may contribute to the climate-driven dengue prediction and dengue risk projection for future climate scenarios.