Relationships between esophageal cancer and spatial environment factors by using Geographic Information System

To explore the relationships between esophageal cancer (EC) and climatic, geographic factors in China by using Geographic Information System, database of EC mortality of 237 sampling areas surveyed in 1990-1992 was established in Excel and linked with the digital polygon maps of study areas. Geographic and climate data of sampling areas were extracted from the raster dataset and finished in Arc/Info 9.0 and ENV14.2 software by using spatial analysis. Spearman correlation analysis and multiple regression analysis after principal component analysis (PCA) were performed to analyze the relationship between EC and these factors. The counties that have the highest EC morality show significant aggregation. Spearman correlation analysis shows weak negative correlation between precipitation, water-heat index (WHI), highest/lowest temperature and EC mortality, and weak positive correlation between drought index (DI), wind speed, population density and EC mortality. Multiple linear regression analysis indicated that the variables associated with EC mortality were precipitation, temperature, wind speed, elevation, DI, WHI and normalized difference vegetation index (NDVI) of July. Our study suggested that the high-risk areas of EC in China are mostly drought and low altitude areas relatively. There were relatively lower NDVI in summer and higher wind speed in these areas. GIS can be applied to cancer epidemiology study and will exert active effect, which should be further explored. (C) 2007 Elsevier B.V. All rights reserved.