A Type 2 Diabetes Risk Predict Model Research Based On Gene And Physical Data

Type 2 diabetes,which is characterized by hyperglycemia,insufficient insulin secretion and insulin resistance,is a worldwide disease.According to the 2014 IFC report,there were 98,400,000 people affected by T2D in China,and the number would climb to 198,400,000 in 2035.Both genetic and environmental factors contribute to the pathogenesis of T2D.But the specific reason is still undercover.With the development of GWAS about T2D,TCF7L2,FTO,KCNQ1 were found to be related to T2D,and discovered part of the heredity of T2D.In many countries,T2D is a main reason of premature death.So the verification of SNP in Chinese and establishing a T2D screening model is necessary.Therefore,we analyzed the susceptible SNP separating capacity in T2D and control.Logistic regression was conducted to verify the relationship between susceptible SNP and T2D in local people(T2D=324,control=700).And we established a risk model based on gene score,the AUC were 0.58(95% confidence interval [CI],0.548-0.622)for equal weighted model,0.583(95% CI,0.546-0.621)for weighted model.At last,a clinical data set(n=5,136)of local population was divided into two groups.Logistic regression on the original data of one group was carried out to screen the risk factor of diabetes.The score of variables was determined based on the ?-coefficients;the aggregate score was used to predict the risk of each person.ROC analysis was used to evaluate the model in another group.The diabetes score value varied from 0 to 114,the AUC was 0.786(95% CI,0.741~0.794).Our results found that the susceptive SNP is meaningful for the prediction of T2D and our result had done some good to the prevention of T2D in China.