摘要
目的在地理信息系统(GIS)及遥感(RS)支持下探讨中国1/10人口抽样地区食管癌死亡率与气候、地理等空间环境因素之间的关系。方法在Excel中建立抽样地区食管癌死亡率数据库(1990-1992年),并与制作的抽样地区县界多边形电子地图连接。在Arc/Info9.0及ENVI4.2中利用空间分析的方法提取抽样地区分县的相应空间环境数据(1961-1990年的年/月平均数据)。对死亡率与相应地区空间环境因素进行Spearman等级相关分析;对环境数据进行因子分析后与死亡率数据进行多元逐步回归分析。结果空间分布地图显示中国食管癌有明显的地区聚集性。等级相关显示男性食管癌死亡率与降水量、水热指数、最高/最低气温呈负相关(r值分别为-0.233、-0.404、-0.143、-0.128;P<0.05或0.01);而与干燥指数,6、8、12月风速,1km2人口数呈正相关(r值分别为0.345、0.189、0.170、0.132、0.140;P<0.05或0.01)。经因子分析与多元逐步回归分析,最后因子4(平均海拔)与因子6(干燥指数、水热指数及7月份NDVI)进入回归方程(标准偏回归系数分别为0.201与0.257,P<0.01)。结论食管癌高发区处于相对干旱的地区,夏季植被覆盖率相对较低,月平均风速偏大,特别是6、8及12月份的平均风速明显大于低发区。GIS/RS技术应用于肿瘤流行病学研究具有重要意义,值得进一步深入探讨。
Objective To explore the relationships between esophageal cancer （EC） with climatic and geographic factors in China using Geographic Information System （GIS） and Remote Sensing （RS）. Methods Database of EC mortality of 237 sampling areas in 1990-1992 was established in Excel and linked with the digital polygon maps of areas. Spatial analysis conducted using geographic and climate data of sampling areas extracted from the raster dataset （ 1961-1990, annual/monthly average data） by Arc/Info 9.0 and ENVI 4.2 software. Spearman correlation analysis as well as multiple regression analysis followed factor analysis was carried out to analyze the relationships between EC mortality and the spatial environment factors. Results Spatial aggregation of EC morality existed significantly in the map of spatial analysis. Spearman correlation analysis showed weak negative correlation between male EC and precipitation, WHI, highest/lowest temperature, r were-0.233, -0.404, -0.143, -0.128 respectively （P〈0.05,P〈0.01）, Whereas factors like DI, wind speed of 6,8,12 months, population per lkm2 found weak positive correlation with male EC, r=0.345, 0.189, 0.170, 0.132, 0.140 respectively（P〈0.05,P〈0.01）. Factor 4 （mean elevation） and factor 6 （DI, WHI and NDVI value of July） entered the multiple regression equation （standard partial regression coefficients were 0.201 and 0.257, P〈0.01 ） after factor analysis and multiple regression analysis. Conclusion EC occurrence related to drought areas, low summer NDVI value and higher monthly wind speeds （especially June, August and December） in areas and GIS/RS method could be applied in cancer epidemiology research.
出处
《华南预防医学》
2008年第1期1-5,9,共6页
South China Journal of Preventive Medicine
关键词
地理信息系统
食管肿瘤
气候
地理
流行病学因素
Geographic information system
Esophageal cancer
Climate
Geography
Epidemiologic factors
