The Potential Role Of Hepatic Function Indexes In Elevated Fasting Blood Glucose And The Risk Of Type 2 Diabetes Mellitus

Objectives:1.Based on gender and fasting blood glucose(FBG)levels,the correlation between various hepatic function indexes and FBG levels in the process of dynamic increase of blood glucose levels were analyzed,and the predictive effect of hepatic function indexes on dynamic increase of FBG levels was discussed.2.Based on gender and age groups,hepatic function indexes and lipids were separately and jointly included to construct a risk prediction model for Type 2 Diabetic Mellitus(T2DM),analyze the differences of risk factors in each group models and compare the potential roles of hepatic function indexes and lipids in predicting the risk of T2DM.Methods:Chapter 1:Potential role of hepatic function indexes in elevated fasting glucose levelsA total of 131944 people who met the relevant inclusion and exclusion criteria for physical examination in the Health Examination Center of the Sixth Affiliated Hospital of Kunming Medical University from January 2014 to March 2018 were selected as the research subjects.The data of gender,age.FBG,hepatic function indexes,lipids and other data were collected for statistical analysis.Subjects were grouped according to gender and FBG levels(<5.0mmol/L,5.0-5.5mmol/L,5.6-6.2mmol/L,6.3-6.9mmol/L,≥7.0mmol/L,respectively).In addition to the general comparison between groups and correlation analysis,multivariate linear regression analysis and multivariate Logistic regression analysis were conducted with FBG level respectively as continuous numerical variable and classified variable(<5.0mmol/L as group 1,5.0-5.5mmol/L as group 2,5.6-6.2mmol/L as group 3,6.3-6.9mmol/L as group 4,≥7.0mmol/L as group 5,respectively)to explore the changes in the correlation between various hepatic function indexes and FBG levels,and to discuss the predictive effect of hepatic function indexes during the dynamic increase of FBG.Chapter 2:Predictive efficacy of hepatic function indexes for T2DM.Total of 121240 subjects included 113,164 people who met the relevant inclusion and exclusion criteria for physical examination in the Health Examination Center of the Sixth Affiliated Hospital of Kunming Medical University from January 2014 to March 2018 were selected as the control group,and 8076 patients with T2DM who met the relevant inclusion and exclusion criteria in the Sixth Affiliated Hospital of Kunming Medical University during the same period were used as the case group.The data of gender,age,hepatic function indexes and lipids of the included subjects were collected for statistical analysis.The subjects were divided into groups according to their gender and age(<25,25～35～,45～,55～,65～,respectively).In addition to the general comparison between groups and correlation analysis,hepatic function indexes and lipids were included separately or jointly for binary Logistic regression analysis,and the stepwise method was used to screen the risk factors of the T2DM risk prediction model in each group,analyze the differences in the composition of risk factors in each group,and compare the potential role of hepatic function indexes and lipids in predicting the risk of T2DM.Results:Chapter 1:Potential role of hepatic function indexes in elevated fasting glucose levels1.Relationship between FBG as continuous variable and hepatic function indexesFor men,there was no correlation between elevated total bilirubin(TBIL)and elevated FBG(P>0.05).Indirect bilirubin(IBIL),Total protein(TP),Albumin(ALB),Globulin(GLO),Alanine aminotransferase(ALT),Aspartate aminotransferase(AST),ALT/AST,Gamma-glutamyltransferase(GGT),Alkaline phosphatase(ALP),α-L-fucosidase(AFU)were positively correlated with FBG level(P<0.05).Direct bilirubin(DBIL)and albumin to globulin ratio(AGR)were negatively correlated with FBG(P<0.05).For women,there was no correlation between TBIL、AGR and FBG(P>0.05).IBIL TP,ALB,GLO ALT,AST,ALT/AST,GGT,ALP,AFU was positively correlated with FBG(P<0.05).DBIL was negatively correlated with FBG(P<0.05).2.The relationship between FBG as categorical variable and hepatic function indexesFor men,TBIL and IBIL of the three bilirubin indexes(TBIL,DBIL and IBIL)had no significance in dynamic prediction of elevated FBG level.The relationship between DBIL and FBG changed during the process of elevated blood glucose,and the potential role of DBIL in predicting dynamic increases in FBG was not yet clear.The significance of serum protein indexes(TP,ALB,GLO,AGR)in each blood glucose group was relatively fixed.The increase of TP,ALB and GLO was positively correlated with the FBG level,while the AGR was negatively correlated.When FBG<6.2mmol/L,serum protein indexes could be used to predict the dynamic increase of FBG level,but with the continuous increase of blood glucose level(≥6.3mmol/L),bilirubin indexes and serum protein indexes were not ideal in predicting the dynamic increase of FBG level.When FBG<5.0mmol/L was used as control,the significance of liver enzymes in each blood glucose group was relatively fixed,and the increase of ALT,AST,ALT/AST,ALP and GGT was positively correlated with the FBG.However,GGT was superior to other liver enzymes in predicting the dynamic increase of FBG level,mainly showing that GGT always maintained a positive correlation with FBG level when the blood glucose level gradually increased.The relationship between AFU and FBG changed during the process of blood glucose increase,and its role in predicting the dynamic increase of FBG level was not clear.For women,when FBG<5.0mmol/L was used as control,TBIL and IBIL of the three bilirubin indexes(TBIL,DBIL and IBIL)was always a risk factor at higher blood glucose levels(5.6-6.2mmol/L,6.3-6.9 mmol/L,≥7.0mmol/L),but it was not ideal for predicting the dynamic increase of FBG.The relationship between DBIL and FBG changed during the dynamic rise of glucose,and its potential role in predicting the dynamic rise of fasting glucose level was not clear yet.When FBG<5.0mmol/L was used as control,the clinical significance of TP,ALB,GLO in other blood glucose groups was relatively fixed,and the increase of TP,ALB,GLO was positively correlated with the FBG.The role of TP and ALB in predicting the dynamic increase of FBG was basically reflected in the FBG<6.2mmol/L,but with the FBG rised further(≥6.3mmol/L),its predictive effect was not ideal.The role of AGR in the dynamic increase of FBG level varied.Combined with the results of linear regression,it was inferred that the change of AGR was not significant during the dynamic increase of FBG in women.ALT,AST,ALT/AST,ALP,GGT was positively correlated with FBG level.ALT/AST,ALP,GGT,AFU were better in predicting the dynamic rise of FBG level,which showed that there was a positive correlation between ALT/AST,ALP,GGT,AFU and the dynamic rise of FBG.Chapter 2:Predictive efficacy of hepatic function indexes for T2DM.1.When hepatic function indexes were included alone to construct the risk prediction model for T2DM,the variable of TP,ALP and AFU were always included in the regression model for all age groups of males.The Odds ratio(ORs)of TP were less than 1,and the ORs corresponding to ALP and AFU were greater than 1.In all age groups of females,TP and AFU were always included in the regression model,and its ORs was similar to that of males.In addition,the ORs corresponding to AFU in all age groups were significantly higher than that of others.The Area under the receiver operating characteristic curve(AUCs)of Receiver operating characteristic curve(ROC)were all greater than 0.9,suggesting that hepatic function indexes had a better predictive effect on T2DM.2.When the risk prediction model of T2DM was constructed by including lipids alone,Apolipoprotein A1(ApoA1)was always included for all age groups of males and females,and the corresponding ORs were all less than 1.The relationship between triglycerides(TG)and T2DM might change with age.When males were younger than 45 years old,the corresponding ORs of lnTG were greater than 1,and the corresponding ORs of lnTG were all less than 1 for males who were older than 54 years old.Logistic regression did not include lnTG in the 45-54 years old.lnTG was included in all female age groups,but its ORs decreased gradually with age.The AUCs were 0.781 to 0.912,and the AUCs varied greatly among gender and age groups.3.When combined hepatic function indexes with lipids to construct the risk prediction model for T2DM,the variable of TP,AFU and ApoA1 were always included for all age groups of males.The ORs of TP and ApoA1 were less than 1 and the ORs corresponding to AFU were greater than 1.Specially,the ORs corresponding to AFU in all groups was significantly higher than others.Logistic regression results of females were similar to those of males.TP,AFU and ApoA1 were included in all age groups.In addition,lnTG was included in all age groups of females,but its ORs decreased gradually with age.4.When separately included in hepatic function indexes,GGT only appears as a risk factor when males were<55 years old and females were>24 years old,and ALP appears as a risk factor for all age groups in males and females younger than 65 years old.However,when combined hepatic function indexes with lipids,the association between ALP,GGT and T2DM risk was no longer statistically significant in several age groups of males and females.Conclusions:1.Liver enzymes(ALT/AST,ALP,GGT,AFU)played a predictive role in the dynamic increase of FBG level,and there were gender differences.It is necessary to discuss the potential role of hepatic function indexes in the elevation of FBG by gender.2.The variables included in the T2DM risk prediction model differed in gender and age groups,regardless of whether the hepatic function indexes and lipids were included separately or in combination.3.The predictive efficacy of hepatic function indexes for T2DM was significantly higher than that of lipids in all age groups of males and females,and the predictive efficacy of the combined application of the both was better.