Predicting dengue incidence in high-risk areas of China through the integration of Southeast Asian and local meteorological factors

Dengue, a climate-sensitive mosquito-borne viral disease, is endemic in many tropical and subtropical areas, with Southeast Asia bearing the highest burden. In China, dengue epidemics are primarily influenced by imported cases from Southeast Asia. By integrating monthly maximum temperature and precipitation from Southeast Asia and local provinces in China, we aim to build models to predict dengue incidence in high-risk areas of China. From 2005-2023, a total of 117,839 dengue cases were reported, with Guangdong and Yunnan provinces accounting for 57.8 % and 26.2 % of cases, respectively. Large outbreaks occurred in 2014 (47,052 cases), 2019 (22,688 cases), and 2023 (19,936 cases), with peak incidence typically observed from August to October. The number of provinces reporting cases and outbreaks gradually linearly increased from 10 and 0 in 2005 to a peak of 30 and 11 in 2019, respectively, before declining in 2021 and 2022, then rebounding to 29 and 8 in 2023. Of the 13,927 imported cases, 91.3 % were from Southeast Asia, primarily from Myanmar (41.3 %) and Cambodia (27.3 %). Predictive models for dengue incidence in Guangdong Province showed high adjusted R² values (0.993-0.999) and deviance explained values (0.975-0.989). The models from Cambodia, Thailand, and the Philippines outperformed the other six models. In Yunnan, adjusted R² values and deviance explained values ranged from 0.81-0.84 and 0.90-0.92, respectively, with models from Laos, Myanmar and Cambodia achieving the best predictive performance. Incorporating meteorological data from Southeast Asia along with local data from China, we were able to develop accurate predictive models for dengue incidence in the high-risk areas of China.