京津唐地区HFMD时空变异分析与影响因子探测

Spatiotemporal Variation Analysis and Risk Determinants of Hand, Foot and Mouth Disease in Beijing-Tianjin-Tangshan, China

手足口病(HFMD)是一种多发于儿童的常见传染病。近年来,中国HFMD发病人数逐年上升,疾病疫情也日益受到社会广泛关注。虽已有不少相关研究,但对于探测其时空异质性及量化潜在影响因子解释力的研究仍然较少。本文采用地理探测器及贝叶斯时空层次模型,对2009-2013年京津唐地区HFMD发病率的时空异质性进行分析,并量化各影响因子及其两两交互作用对HFMD发病率的解释力。结果表明:①HFMD的相对风险存在时间异质性,其发病风险在春夏季(5-7月)达到峰值,而冬季(12-次年2月)发病风险最低;②HFMD的相对风险存在空间异质性,在经济发达的地区HFMD发病率较高;③影响HFMD发病率时间异质性的主要气象因子为平均温度、累积降水、相对湿度,解释力分别为0.38,0.27,0.13,且交互作用都大于独自影响的作用,如平均温度和相对湿度、平均温度和降水,平均温度和风速两两交互的解释力分别为0.43, 0.40, 0.42。通过研究京津唐地区HFMD发病率的时空异质性以及影响因子的量化状况,为本地区HFMD的预防和控制提供理论依据。

Hand, foot and mouth disease (HFMD) is a common infectious childhood disease. In recent years, the number of cases of HFMD in China has increased rapidly, and has received increasing attention. Although there are many related studies, only a few studies focus on the spatiotemporal heterogeneity of HFMD incidence and quantify the association between meteorological factors, socioeconomic variables, and HFMD incidence. Geodetector and Bayesian space-time hierarchical models were applied to analyze the spatiotemporal heterogeneity of the HFMD incidence from 2009 to 2013 within the Beijing-Tianjin?Tangshan region. These were used to quantify the determinant power of meteorological factors, socioeconomic variables, and the interactions between two of these factors. The Geodetector method has the axiom that if an explanatory variable (x) detennined an explained variable (y), the explained variable would exhibit a spatial distribution similar to that of the explanatory variable. This method has been widely used to measure the detenninant power of potential explanatory variables. The Bayesian space-time hierarchical model has the potential to show the spatiotemporal variation of a geographic phenomenon. The results showed that:(1) the highest incidence of HFMD occurred in late spring and summer (May to July), and the lowest incidence occurred in winter (December to February).(2) Spatial heterogeneity existed. In particular high risks areas were mainly concentrated in areas of high economic development. The population density and proportion of the tertiary industry determinants, play a lead role in contributing to the spatial heterogeneity of HFMD incidence (q values of 0.35 and 0.2& respectively, as calculated by GeoDetector).(3) The main meteorological factors affecting the temporal heterogeneity of HFMD incidence were average temperature, cumulative precipitation, and relative humidity (with a determinant powers calculated by GeoDetector of 0.38, 0.27 and 0.13, respectively). Additionally, the interactions were greater than the independent effects between socioeconomic variables or meteorological factors. For example, the interaction of average temperature and relative humidity, average temperature and precipitation, average temperature and wind speed were 0.43, 0.40 and 0.42, respectively. The interaction of population density and proportion of the tertiary industry was 0.55. This result presented the strongest correlation with HFMD incidence. Temperature and relative humidity were also dominant factors influencing the spatiotemporal transmission of HFMD, along with areas of high economic development with high population density. This study provides a theoretical basis for the prevention and control of HFMD by detecting the spatiotemporal heterogeneity of the HFMD incidence and quantifying the impact factors within the study region.