Analysis Of The Influencing Factors Of Dengue Fever And Its Prediction In Typical Epidemic Regions In China

Dengue fever(DF)is a common mosquito-borne viral infectious disease in the world,and increasingly severe DF epidemics in China have seriously affected people’s health in recent years.Thus,investigating spatiotemporal patterns,transmission regularity and potential influencing factors of DF epidemics in different spatial scales of typical regions can provide an important reference for local disease prevention and control departments to formulate solutions and reduce risks.In addition,predicting the high-risk areas of dengue fever epidemics in the future is critical to consolidate effective prevention and control measures for these regional epidemics.From the perspective of comparison,the generalized additive models(GAM)and land use regression models(LUR)were used in our study to identify and compare potential contributing factors that influence spatiotemporal epidemic patterns in typical DF epidemic regions of China(e.g.,the Pearl River Delta [PRD] and the Border of Yunnan and Myanmar [BYM]),and to analyze the consistency and difference of model results.In terms of influencing factors,environmental factors including the normalized difference vegetation index(NDVI),temperature,precipitation,and humidity,in conjunction with socioeconomic factors,such as population density,road density,land-use,and gross domestic product,were employed.Finally,taking the Pearl River Delta region as an example,based on the 2030 land use data simulated by the SLEUTH model and the 2030 population density data predicted by the population prediction model,the spatial distribution characteristics of the 2030 dengue risk are predicted according to the LUR multiple regression equation.The main findings and conclusions of this study are as follows:(1)DF epidemics in the PRD and BYM exhibit prominent spatial variations,which are characterized by significant spatial clustering over the Guangzhou-Foshan,Dehong,and Xishuangbanna areas.(2)The GAM that integrated the population density-urban land ratio-NDVI-humidity-temperature factors for the PRD and the urban land ratio-road density-NDVI-temperaturewater land ratio-precipitation factors for the BYM performed well in terms of overall accuracy,with a total variance of 83.4% and 97.3%,respectively.The results show that the environmental and socioeconomic factors in the PRD and BYM may affect the spatialtemporal differentiation of DF epidemics.However,socioeconomic factors have a stronger influence on DF epidemics than environmental factors in both areas.As indicated,the differences in socioeconomic factors are more obvious in cases where environmental factors are suitable and differ slightly throughout areas,and the influencing mechanisms have their own regional characteristics.Among these socioeconomic factors,the PRD is mainly affected by population density and urban land ratio,and the BYM is mainly affected by urban land ratio,road density and water area ratio.Whereas NDVI is the main environmental factor in both regions.In addition,the common factors(urban land ratio,NDVI,and temperature)in these two regions exhibited different effects on regional epidemics.The LUR that integrated the population density-urban land ratio-grass land ratio-forest land ratio-farmland ratio factors for the PRD and the village area ratio-water land ratio-urban land ratio-relative humidity-road density factors for the BYM performed well with the fitness between measured and fitting values of 0.7101 and 0.6082,explaining a total variance of 79.6% and 83.2%,respectively.The socioeconomic factors still have a stronger influence on DF epidemics than environmental factors in both areas.Among these socioeconomic factors,the areas with high population density,high urban land ratio and certain vegetation coverage will increase the risk of DF in the PRD.As for BYM,dengue epidemic risk will increase in areas with relatively good road network conditions such as towns and rural areas,while the epidemic tends to be higher in areas around water and with certain humidity conditions.In the environmental conditions,no factors enter the LUR model of PRD,while only the average relative humidity factor enters the LUR model in BYM.(3)Both GAM and LUR models have their own characteristics and limitations in our study.The GAM model shows a better fitting effect by predicting non-linear factors more accurately,but it will be affected by abnormal fluctuations of data.The LUR linear regression model has some limitations in analyzing the spatial distribution and influencing factors of dengue epidemic in reality,but it is simpler and more efficient in simulating the risk of dengue epidemic based on socioeconomic factors.In addition,compared with the BYM,the LUR model in the PRD has a higher fitness,which is more feasible for future dengue risk prediction.(4)In 2030,the spatial distribution of dengue risk in the PRD will be consistent with the current highly urbanized and highly populated areas,which are clustering over Guangzhou-Foshan area,as well as the northern part of Dongguan and Shenzhen.The aggregation of dengue fever risk will be weakened,but with the expansion of urbanization,the dengue epidemic has a tendency to spread to highly urbanized areas.These results of dengue risk prediction can provide relevant departments with high risk areas of dengue fever that need to be focused on in the future.