[Application of different spatial interpolation methods in sodium intake estimation].

Abstract

OBJECTIVE\nThe spatial interpolation method was used to estimate the intake of sodium in the areas without dietary sodium intake data.\n\n\nMETHODS\nThe data of dietary sodium consumption in this study are from the 2010-2012 China National Nutrition and Health Surveillance(CNNHS), the CNNHS was carried out on stratified multistage systematic clustered random sampling method with proportional to the population to form a representative sample of China as a whole, including 150 survey site. Dietary data was collected through face-to-face interview, based on a combination of three consecutive days of 24-h recalls combined with a household food weighting method, a total of 48 826 participants aged 20 years old and more were recruited in this study. The average sodium intake among 150 survey sites were calculated according to data of the CNNHS 2010-2012. The accuracy of spatial analysis techniques, such as Inverse Distance Weight, Ordinary Kriging, Cooperative Kriging method and Regression Kriging Interpolation method, was compared in the estimation of dietary sodium intake. The spatial analysis method with the highest accuracy was used to estimate the daily dietary sodium intake of each district and county in China, and then the dietary sodium intake of each province was calculated according to the population weight.\n\n\nRESULTS\nThe average sodium intake among Chinese adults was 5. 18 g, among which the average intake of men is 5. 33 g and that of women was 5. 03 g. Comparing the accuracy of four spatial analysis method in estimating dietary sodium intake, it was found that the Regression Kriging Interpolation method was superior to the other three methods, with root mean square error and mean absolute error of 0. 54 and 0. 44. The Regression Kriging Interpolation method was used to estimate the dietary sodium intake in different regions of China, the estimated national dietary sodium intake by population weighted calculation was 5. 17 g/d, the estimated national dietary sodium intake was 5. 33 g/d for men and 5. 01 g/d for women. The difference rate between the measured and estimated dietary sodium intake for all age groups was not more than 10%. Shandong Province, Tianjin City and Hebei Province had the highest sodium intake, which was 5. 98 g/d, 5. 83 g/d and 5. 82 g/d, respectively. Guizhou Province had the lowest sodium intake, with an estimated intake value of 4. 27 g/d.\n\n\nCONCLUSION\nThe spatial interpolation method can be used to estimate the intake of sodium, and the accuracy of Regression Kriging Interpolation method is higher than other spatial interpolation method.