关键词: 血尿素氮, 地理因素, 主成分分析, 岭回归分析, 回归分析

Abstract: Objective: To discuss the relationship between the geographic environment and blood urea nitrogen(BUN) reference values of the healthy people, and to explore the distributional rule of BUN reference values of the healthy people, and to provide the scientific foundation for establishing the BUN reference value standards of different areas. Methods: A total of 23 geographic factors and 33521 BUN reference values of healthy adults measured by 403 medical facilities from 23 provinces, 4 municipalities and 5 autonomous regions were collected. The spatial autocorrelation analysis was used to determine the spatial autocorrelation of the sample data; the correlation analysis was used to detect the factors which correlated significantly with the BUN reference values;the multiple linear regression, principle component analysis and ridge regression analysis were respectively used to construct the predicted models; the paired-sample t test was used to choose the optimal model; the distribution map of BUN reference values was built by geostatistic analysis. Results: There were 5 geographic factors, latitude(X₂), altitude(X₃), annual mean temperature(X₅), annual mean relative humidity(X₆) and annual precipitation(X₇), correlated significantly with the BUN reference values. The regression equation of optimal model was ?=5.112+0.000 127 1X₂+0.000 094 61X₃-0.000 140 5X₅-0.000 136 8X₆-0.000 139 1X₇±0.531 0; the distribution map of predicted values of the BUN reference values was obtained. Conclusion: The overall trend of BUN reference values is low in the east and high in the west. The BUN reference value is negatively associated with the altitude. If the geographic data of a certain region could be obtained, the BUN reference value of this region will be predicted.

Key words: geographical factors, principal component analysis, blood urea nitrogen, ridge regression analysis