Prospective Study On The Prediction Of Diabetes Risk By Obesity Evaluation Indexes And Their Optimal Cut-off Values

Objectives To compare the ability of BMI,WC and WHtR to predict the risk of diabetes,to compare the ability of single evaluation index and joint evaluation index to predict the risk of diabetes,to explore the optimal cut-off value of obesity evaluation index in predicting the risk of diabetes,and to analyze and verify the impact of dynamic changes of obesity on the risk of diabetes.Methods Based on the jinchang cohort platform,this study adopted prospective research methods to include 10,895 subjects?6,317 males and 4,578 females?.Cox proportional risk model was used to analyze the risk of diabetes at different levels of body mass index?BMI?,waist circumference?WC?and waist to height ratio?WHtR?,Including univariate model,model 1 and model 2:model 1 adjustment variables are age,smoking and alcohol consumption history,family history of diabetes,occupational history,and educational level),model 2 adjustment variables are model 1adjustment variables plus SBP,TG/HDL.The linear regression model was used to explore the relationship between BMI,WC,WHtR and fasting blood glucose.Adjustment variables are age,smoking and alcohol consumption history,family history of diabetes,occupational history.Draw ROC curve to calculate the area under the curve of obesity evaluation index and joint index,compare AUC to analyze the prediction ability of obesity evaluation index and joint index to diabetes,and obtain the optimal cutoff value through the maximum point of Youden index.The model 1independent variable is the index of obesity evaluation,the model 2 independent variable is the model 1 independent variable and age.The interaction of BMI,WC and WHtR was analyzed by multiplying model.Adjustment variables are age,smoking and alcohol consumption history,family history of diabetes,occupational history.Logistic regression model was used to analyze the dynamic change of obesity evaluation index and the correlation between the dynamic change of overweight or obesity status and the incidence of diabetes.The adjustment variables of model 1 were age,baseline fasting blood glucose,baseline obesity evaluation index values.The adjustment variables of model 2 were the adjustment variables of model 1 plus smoking and alcohol consumption history,family history of diabetes,occupational history,and educational level.Results 1.In men,women and the total population,the incidence of diabetes increased with the increase of BMI,WC and WHtR?P<0.05?.2.In the univariate model,model 1 and model 2 of Cox proportional risk regression analysis,the risk of diabetes with different BMI,WC and WHtR levels in both men and women gradually increased with the increase of BMI,WC and WHtR,and all of them showed an increasing trend?P<0.05?.In the linear regression model,BMI,WC and WHtR were significantly correlated with fasting blood glucose in both men and women?P<0.05?.3.Pearson correlation showed that the trend of BMI,WC and WHtR was similar in men and women,and the correlation between BMI,WC and WHtR was statistically significant?P<0.05?.In men and women,BMI and WC,BMI and WHtR had a forward multiplying interaction on the incidence of diabetes(P_interaction<0.05),and BMI and WC,BMI andWHtR had a synergistic effect,which increased the risk of diabetes.4.Among the total population,male and female subjects,both model 1 and model 2 had the maximum AUC for BMI and the minimum AUC for WC.In the female and total population of model 1,WC and BMI AUC checked by Z,the difference was statistically significant?P<0.05?.There was no statistically significant difference in the AUC of male BMI with WC and BMI with WHtR?P>0.05?.In model 2,there were statistically significant differences in the AUC of BMI with WC,BMI with WHtR among males,females and the total population?P<0.05?.There was no statistically significant difference in the AUC of male BMI with WC and BMI with WHtR?P>0.05?.In model 2,there were statistically significant differences in the AUC of BMI with WC,BMI with WHtR among males,females and the total population?P<0.05?.When the combination of BMI and central obesity indicators WC and WHtR was used to predict diabetes,both of them were greater than the AUC predicted by a single obesity evaluation index.5.The optimal cut-off value of BMI in predicting diabetes was 23.4 kg/m² in the total population,24.6 kg/m² in men,and 23.4 kg/m² in women.The optimal cut-off value of WC was 85.5cm in the total population,89.5cm in men and 76.5cm in women.The optimal cutoff for predicting diabetes was 0.52 in the total WHtR population,and 0.52 in men and 0.47 in women.6.In both male and female subjects,except for male WHtR in model 1?adjusted for age,baseline fasting blood glucose and baseline obesity evaluation indexes?and model 2?adjusted for variables in model 1 plus the smoking,alcohol consumption history,occupational history,educational level and family history of diabetes?,the risk of Q4 diabetes in BMI,WC and WHtR quartile group was significantly different from that in Q1?P<0.05?,and the risk of diabetes was gradually increasing(P_trend<0.05).Compared with the baseline survey in 2013,the diabetes risk of those who changed from normal weight to overweight and obesity was increased,and the diabetes risk of those who changed from overweight and obesity to normal weight was decreased.Conclusions 1.BMI,WC and WHtR are risk factors for diabetes,and there is a linear correlation and interaction between BMI,WC and WHtR,which together increase the risk of diabetes.2.BMI and WHtR have strong ability to predict diabetes,and the combination of BMI with WC and WHtR can improve the ability to predict the incidence of diabetes.3.The optimal cut-off values of obesity evaluation indexes are different between men and women.4.The optimal cut-off values of BMI and WC was lower than the current standard.It was suggested that the optimal cut-off value of BMI was 23 kg/m².5.Weight control can reduce the risk of diabetes.