Analysis Of Clinical Features And Establishment Of Prediction Model Of Latent Autoimmune Diabetes In Adults

Objective:To investigate the prevalence of latent autoimmune diabetes in adults(LADA)in newly diagnosed type 2 diabetes mellitus(T2DM)in Tianjin.To analyze the similarities and differences of clinical features and pancreatic beta cell function between LADA patients and T2 DM patients.To establish and validate prediction model for LADA in newly diagnosed T2 DM patients.Methods:Outpatients aged 18 and older who were diagnosed with diabetes mellitus(DM)within one year were collected in Tianjin Medical University Metabolic Diseases Hospital Identification Center of diabetes outpatient specific diseases during May,2015 to April,2018.General information,diabetes status,other disease history,family history of chronic diseases and personal history of the patients were collected through questionnaires.Measure and record height,weight,waist circumference,hip circumference,calculate body mass index(BMI)and waist hip ratio(WHR).Measure glycosylated hemoglobin A1c(HbA1c),perform 75 g oral glucose tolerance test(OGTT),insulin releasing test(IRT),and C-peptide releasing test(CRT).Record the results of blood glucose,insulin and C-peptide at different time points.According to the approximate trapezoid area formula,the area under the curve(AUC)corresponding to PG,INS and CP at two hours,the first hour and the second hour were calculated respectively.The HOMA2 calculator was used to calculate the homeostasis model assessment of insulin resistance index(HOMA2-IR)and pancreatic beta cell function(HOMA2-β).Serum glutamic acid decarboxylase antibody(GADA),insulinoma-associated antigen-2 antibody(IA-2A)and zinc transporter 8 antibody(ZnT8A)were detected by enzyme-linked immunosorbent assay(ELISA).According to the classification standard of glucose metabolism and the diagnostic standard of diabetes established by the World Health Organization(WHO),DM patients were screened out,and those patients with Gestational Diabetes Mellitus(GDM)and suspected or diagnosed with secondary diabetes were excluded.According to the diagnostic criteria proposed by the consensus of diagnosis and treatment of LADA formulated by Chinese Diabetes Society(CDS),LADA patients were screened out.Calculate the prevalence of LADA in newly diagnosed T2 DM patients,and analyze the differences and similarities in clinical features and pancreatic beta cell function between the two groups.In a ratio of 2:1,the patients were divided into two groups: the modeling group and the verifying group The predictors related to LADA diagnosis in T2 DM were screened out by univariate logistic regression analysis,and then the regression model was established by multivariate logistic regression analysis.Hosmer-Lemeshow was used to test the fitting degree of the evaluation model.The area under the receiver operating characteristic(ROC)curve of prediction probability was used to evaluate the discrimination of the model by the verifying population.The Youden index was calculated according to the sensitivity and specificity of the ROC curve coordinate points,and the cut-off point of the model predictive value was determined by combining the statistical index with the clinical situation.Results:Finally,930 patients were included in this study,including 880 patients(94.6%)with T2 DM and 50 patients(5.4%)with LADA.Age,proportion of abdominal obesity,1hINS,2hINS,FCP,1hCP,2hCP,HOMA2-β of patients in LADA group were lower than those in T2 DM group,incidence of ketosis,proportion of smoking history,HbA1 c,FPG,1hPG and 2hPG were higher than those in T2 DM group,and the differences were statistically significant(P<0.05).Univariate logistic regression analysis showed that age,ketosis,smoking history,HbA1 c,FPG,1hPG,2hPG,1hCP,2hCP,PG-AUC,1hPG-AUC,2hPG-AUC,CP-AUC,1hCP-AUC,2hCP-AUC,HOMA2-β were independent predictors of LADA in newly diagnosed T2 DM patients(P<0.05).Multivariate logistic regression equation after variable screening was as follows:Y =-8.249-0.035X1 + 1.755X2 + 1.008X3 + 0.321X4-0.126X5,where Y is DM(0=T2DM,1=LADA),X1 is age,X2 is ketosis(1=NO,2= YES),X3 is smoking history(1=NO,2=YES),X4 is 1hPG-AUC,X5 is 2hCP-AUC.The Hosmer-Lemeshow test showed that the model could fit the data(P>0.05),and the area under the ROC curve of the prediction probability of the verifying population was 0.757,suggesting that the model had a better discrimination(P<0.05).When Y=0.0472,the sensitivity and specificity of the model are 78.57% and 67.96% respectively.At this time,the youden index is the largest and the value is 0.47.Conclusion:1.In Tianjin,5.4% of newly diagnosed T2 DM patients are actually LADA patients.LADA patients and T2 DM patients have no difference in gender,ethnicity and education level,and there is no significant difference in the "three more and one less" symptoms and the proportion of patients with hypertension,coronary heart disease and hyperlipidemia.Age and abdominal obesity are significantly lower in LADA patients than in T2 DM patients,and smoking history is significantly higher in LADA patients than in T2 DM patients.2.Compared with patients with T2 DM,patients with LADA had obvious decline in the function of pancreatic beta cell at the beginning of diagnosis,and had less remaining pancreatic beta cells and insulin secretion,which was also reflected in the fact that patients with LADA were more prone to ketosis,and had high HbA1 c and poor blood glucose control.3.There was no significant difference in the degree of insulin resistance between patients with LADA and those with T2 DM,indicating that the occurrence of LADA is not only the destruction of pancreatic cells caused by autoimmunity,but also the role of insulin resistance.4.For newly diagnosed T2 DM patients,this model was used for calculation.When Y≥0.0472,combined antibody screening was recommended.When Y<0.0472,according to the patient’s economic situation and personal willingness to decide whether to screen antibody.