首页|基于MGWR模型的西安手足口病发病影响因素

基于MGWR模型的西安手足口病发病影响因素

扫码查看
目的 在 5 km×5 km的空间网格尺度上探讨西安市手足口病发病与环境、社会经济因素等的关系,为区域防控措施的制定提供依据.方法 收集 2019 年西安市手足口病报告发病率数据,应用空间自相关分析手足口病空间分布特征;基于多尺度地理加权回归(multiscale geographically weighted regression,MGWR)模型分析环境与社会经济因素对手足口病发病的影响,并与普通最小二乘(ordinary least squares,OLS)回归模型以及地理加权回归(geographically weighted regression,GWR)模型结果进行对比.结果 2019 年西安市手足口病年报告发病率为157.99/10 万,在空间分布上存在正相关性(全局Moran's I=0.349,P<0.001).MGWR模型拟合度优于GWR模型和OLS模型(MGWR:R2=0.530;GWR:R2=0.473;OLS:R2=0.327).各影响因素的作用尺度存在一定差异,GDP、土地城镇化水平、平均气温等作用尺度较大,归一化植被指数(normalized difference vegetation index,NDVI)作用尺度较小.GDP与手足口病报告发病率呈显著负相关,土地城镇化水平、平均气温与报告发病率呈显著正相关,NDVI在西安部分地区对手足口发病有显著负向影响.结论 环境与社会经济因素对手足口病发病有显著影响,且各影响因素的作用存在空间差异,研究结果可为不同地区制定针对性的预防措施提供依据.
Influencing factors on the incidence of hand,foot and mouth disease in Xi'an based on MGWR model
Objective To investigate the relationship between the incidence of hand,foot and mouth disease(HFMD)and factors related to environment and socioeconomic in Xi'an at a spatial grid scale of 5 km,and provide a basis for the development of regional control and prevention measures.Methods The 2019 HFMD report data in Xi'an were collect-ed and analyzed by spatial autocorrelation to characterize the spatial distribution.The role of natural environmental fac-tors and socioeconomic factors on the incidence of HFMD was analyzed based on a multiscale geographically weighted regression(MGWR)model and compared with those of the ordinary least square(OLS)regression model and the geo-graphically weighted regression(GWR)model.Results The annual reported incidence rate of HFMD in Xi'an in 2019 was 157.99/100,000,with a positive correlation in spatial distribution(global Moran's I=0.349,P<0.001).The MGWR model fit was better than the GWR model and the OLS model(MGWR:R2=0.530;GWR:R2=0.473;OLS:R2=0.327).On the scale of influence,the GDP,land urbanization level,and average temperature had larger scale effects,while normalized difference vegetation index(NDVI)had smaller scale influence.GDP was significantly nega-tively correlated with the reported incidence of HFMD,land urbanization level and average temperature were significant-ly positively correlated with the reported incidence,and NDVI had a significant negative effect on HFMD incidence in parts of Xi'an.Conclusion The influence of environmental and socioeconomic factors on the incidence of HFMD is significant and there are spatial differences in the role of each influencing factor.The results are helpful for the formula-tion of regional prevention and control strategies for HFMD.

Hand,foot and mouth diseaseSpatial autocorrelationMulti-scale geographically weighted regressionGeographically weighted regressionXi'an

梁珂梦、李树芬、倪志松、宋思豪、席睿、程传龙、左慧、段雨琪、刘昆、白尧、李秀君

展开 >

山东大学齐鲁医学院公共卫生学院生物统计学系,山东 济南 250012

空军军医大学军事预防医学系军队防疫与流行病学教研室,陕西 西安 710032

西安市疾病预防控制中心传染病预防控制科,陕西 西安 710054

手足口病 空间自相关 多尺度地理加权回归 地理加权回归 西安

国家重点研发计划

2023YFC2604400

2024

山东大学学报(医学版)
山东大学

山东大学学报(医学版)

CSTPCD北大核心
影响因子:0.841
ISSN:1671-7554
年,卷(期):2024.62(6)
  • 15