| At present,chronic non-communicable diseases(chronic diseases)have become one of the major threats to human health worldwide,among them,cancer,cardiovascular disease,diabetes and respiratory diseases account for 82%of the total mortality of chronic diseases.According to the data,the nu:ber of chronic disease patients in china has exceeded 300 million,and the lethality rate accounts for 85%of the total number of deaths in china every year.Chronic diseases have become one of the serious financial burdens in china due to their long course and high cost of treatment.Therefore,early screening and prevention of chronic diseases are particularly important.Serum adiponectin(APN)and fasting insulin(FINS)levels are blood indicators for evaluating diabetes and cardiovascular diseases.APN levels can indicate the occurrence and development of metabolic diseases such as type 2 diabetes and atherosclerotic cardiovascular and cerebrovascular diseases,FINS is an early diagnostic marker of diabetic nephropathy.Current studies involve the study of APN and FINS in terms of age,gender,ethnicity and plateau environment,however,there is a lack of research on the impact of other geographic environmental factors on APN and FINS.APN and FINS are affected by both genes and environment.When making medical reference values,the influence of geographical environment on them is ofton neglected.Therefore,according to the research ideas and methods of geography,this paper explores the relationship and distribution between APN and FINS and geographical factor.This paper attempts to provide methods and basis for the formlation of reference values ofmedical indicators,improve the formulation of standards and increase the scientificity.The APN reference values of 13988 healthy person frol 214 hospitals and related medical units in china and the FINS reference values of 12627 healthy person from 108 hospitals and related medical units in china were collected from 2008 to 2018.A total of 12 geographical factors were selected for analysis,including longitude,latitude,altitude.annual sunshine hours,annual mean temperature,annual precipitation,annual temperature difference,annual average relative humidity,topsoil organic content,topsoil pH,total exchangeable amount of topsoil and topsoil salinity.Through spatial autocorrelation analysis,the spatial correlation of medical indicators was explored to determine the feasibility of the study.Based on the geographic detectors,the factors that have the greatest impact on the spatial distribution of medical indicators are determined.According to the results of collinearity diagnosis,appropriate modeling methods are selected for analysis.In this paper,ridge regression,principal component analysis and support vector machine are selected to model and build a combined model.Different model methods are evaluated and the best prediction model is selected to predict,and prediction values of 2322 data points in China are obtained.Geostatistical analysis method was used to analyze the predicted data,and appropriate interpolation method was used to draw the spatial distribution maps of the reference values of two medical indicators,and observe the spatial distribution law.According to the results of spatial autocorrelation and correlation analysis,there is spatial correlation between the two medical indicators.APN medical indicators are correlated with six geographic factors:latitude,altitude,annual mean temperature,annual precipitation,topsoil pH,total exchangeable amount of topsoil.FINS medical indicators are correlated with five geographic factors:longitude,altitude,annual mean temperature,annual precipitation,annual average relative humidity.The results of geographic detectors show that altitude is the most important factor affecting the spatial distribution of the two medical indicators.Because of the collinearity problem,ridge regression,principal component analysis and support vector machine are selected to model and a combination model is established based on a single model.Through the test,the optimal prediction models of the two medical indicators are the combination model of reciprocal variance method.According to the spatial distribution maps of the two medical indicators,it can be concluded that the low reference values of APN and FINS in Chinese adults are distributed in the Tibet Plateau and its surrounding areas.This paper mainly studies the relationship between APN and FINS medical indicators reference values and geographical factors.Based on medical indicators and geographical factors.aims to find out the spatial distribution of reference values and try to explain them with geographical principles.Using the research method of geography,the optimal prediction model of APN is:YAPN??=0.332Y ?+0.317Y?+0.351Y?,the optimal prediction model of FINS is:YFINS??=0.322 Y?+0.319Y?+0.359Y? To provide reference standard for formulating more accurate and scientific reference values of medical indicators.Incorporating geographical factors into the reference criteria can make the criteria suitable for the local population according to local conditions,which is conducive to improving the scientificity of clinical diagnosis. |
PDF Full Text Request