Spatial characteristics of non-iodized salt at household level based on geographic information system in Chongqing / 中国地方病学杂志

ABSTRACT

Objective To analyze the spatial distribution characteristics of non-iodized salt at household level based on geographic information system (GIS) in Chongqing.Methods The database of non-iodized salt at county level from 2001 to 2010 was established in Chongqing.By using GIS technology,the spatial distribution and spatial autocorrelation were analyzed by ArcGIS 9.3 software.Results The rate of non-iodized salt was fluctuated between 2.35％-5.78％ during 2001-2006 and the rate was reduced to less than 2.00％ after 2007.The result of spatial autocorrelation analysis on non-iodized salt from 2001 to 2006 indicated that Moran's Ⅰindex was 0.145578,0.078801,0.108033,0.091957,0.127749,0.214302,respectively(Z value was 3.066275,1.977321,2.541619,2.309972,2.900446,3.874203,respectively,all P ＜ 0.05).The spatial distribution of non-iodized salt had marked spatial cluster through Chongqing region from 2001 to 2006.The result of local spatial autocorrelation analysis from 2001 to 2006 indicated that Fengdu and Fuling were two high-risk areas(all P ＜ 0.05).Dianjiang,Yubei,Jiangbei,Wulong and Banan were also confirmed as high-risk areas in 2001,2005 and 2006(all P ＜ 0.05).The results also indicated that the distribution of non-iodized salt in the seven high-risk areas was positively correlated.The result of spatial autocorrelation analysis on non-iodized salt from 2007 to 2010 indicated that Moran's Ⅰ index was 0.018361,0.016186,0.040769,-0.059691,respectively (Z value was 1.093310,0.787361,1.071811,-0.583820,respectively,all P ＞ 0.05).The spatial distribution of non-iodized salt was at random on the whole from 2007 to 2010.However,there were four local high-risk areas.The distribution in Fengdu and Dianjiang was positively correlated,while that in Jiangjin and Shizhu was negatively correlated.Conclusions The distribution of non-iodized salt at households level in Chongqing is changed from spatial distribution before 2006 to random distribution after 2007,but there are high value areas,which should be taken as the focus of monitoring.