Association Study Between Anthropometric Measures And β-cell Function In Obese Patients With Different Glucose Metabolism

Obesity is an important risk factor for glucose metabolism disorders.With the change of glucose metabolism from normal to abnormal,insulin secretion in obese patients compensatory increased in the early stage,and a slow decline in prediabetes,only 50%of normal levels in diabetes stage.Body measurement index BMI,WC and WHR are common indicators of obesity.The reports on the relationship between beta cell function（BCF）and anthropometric index in obese patients are inconsistent,and most of the studies are based on normal glucose metabolism,but the relationship between anthropometric index and BCF in obese people under different states of glucose metabolism has not been reported.There are many methods to evaluate BCF,among which oral glucose tolerance test（OGTT）is a common method to evaluate BCF,which is relatively simple and the first choice for large-scale epidemiological studies.There are many kinds of BCF index based on OGTT,and the choice of index is different in existing studies,and which index is most suitable for obese people has not been reported.Based on the above background,this study first evaluated 19 OGTT-based BCF indexes to screen the most suitable indexes for evaluating BCF in obese patients.Then,the correlation between anthropometric index and BCF was studied under different glucose metabolism states,and the index that best reflects BCF in obese patients with different glucose metabolism was selected.The relationship between body fat distribution index,anthropometric index and BCF measured by Dual energy x-Ray absorption（DEXA）was analyzed to verify the optimal index.Finally,the relationship between the change of BCF and anthropometric index after weight loss was studied,and the best index was used to construct the BCF prediction model.Part I Selection on OGTT basedβ-cell function index in obese patients with different glucose metabolismObjective:To screen out a simple index that can better evaluate the BCF of obese people among 19 OGTT-based BCF indices.Methods:847 obese patients who came to the hospital for the first time were divided into three groups according to the status of glucose metabolism:normal glucose metabolism group（n=427）,prediabetes group（n=248）and type 2 diabetes group（n=172）.The insulin sensitivity index（ISI）and the 19 BCF indexes reported in the literature were calculated using the results of OGTT.Theoretically,the relationship between BCF index and ISI is hyperbolic,from which a new index,disposal index DI,can be calculated.ROC AUC method was used to evaluate the ability of these BCF index and DI index to distinguish between prediabetes and type 2 diabetes,and the indexes with higher differentiation power were screened out.Results:Through screening,we found that among the 19 commonly used BCF indices,CIR₃₀ was the first discriminant（ROC AUC:90.6,95%CI:88.1～93.3）.93.4,95%CI:90.8-96.1)can play a better role in distinguishing between T2DM and non-diabetic groups and between T2DM and NGM groups.in addition,CIR₃₀ ranks second in distinguishing between NGM and pre-diabetic groups（ROC AUC:70.9%,95%CI:66.9to 74.9）.Other indices such asΔI₃₀/ΔG₃₀,BIGTT_0.30.120,CIR₁₂₀,ΔI₆₀/ΔG₆₀,AUCI_60-120/G_60-120 also have higher distinguishing power（ROC AUCs>80%）.In NGM and prediabetes group,I₁₂₀/I₀,CIR₃₀,ΔI_60–120/ΔG_60–120 index had better distinguishing power（ROC AUCs>70%）.In the same type of index(such as CIR₃₀,CIR₁₂₀),the discriminant ability of early phase BCF index is better than that of late phase BCF index.HOMA-β,CIR₃₀,ΔI₃₀/ΔG₃₀ and AUCI_60-120/G_60-120 showed hyperbolic fitting with ISI and had the best degree of fit.After calculating DIs,CIR₃₀ and AUCI_60-120/G_60-120 have better discriminant ability.Conclusion:Among the obese people in this study,CIR₃₀ showed the best ability to distinguish the state of glucose metabolism and had a high degree of fit with the hyperbolic relationship of ISI.Therefore,CIR₃₀ is a reliable and convenient evaluation index when evaluating BCF in the following study.Part II Cross-sectional study on the relationship between anthropometric index andβ-cell function in obese patients with different glucose metabolismObjective:To study the relationship between BMI,WC,WHR and BCF index in different states of glucose metabolism,and whether these relationships are independent of the effect of insulin resistance on BCF.To find out the index of BCF which can best reflect the different glucose metabolism of obese patients.Methods:847 obese patients,including normal glucose metabolism group（n=427）,prediabetes group（n=248）and type 2 diabetes group（n=172）.The main calculations were carried out in three groups.BMI,WC and WHR were measured in each patient,and multiple body fat data were measured by dual energy X-ray absorptiometry（DEXA）.In the study,CIR₃₀ and CIR₃₀ DI were used to evaluate BCF and HOMA-IR was used to assess insulin resistance.Pearson correlation analysis was used to evaluate the correlation among BMI,WC,WHR and BCF index;multiple stepwise regression analysis was used to determine the best independent body measurement index for predicting BCF index;intermediary analysis was used to verify the mediating effect of insulin resistance in obesity-induced compensatory changes in BCF.Finally,the correlation between anthropometric index andβ-cell function was verified by analyzing the correlation between more accurate body fat measurement index and BCF index.Results:In the different glucose metabolic states from NGM to prediabetes and then to T2DM,BMI was always positively correlated with CIR₃₀,WC was also positively correlated with CIR₃₀（no statistical significance）,for WHR,it was positively correlated with CIR₃₀ in NGM group（r=0.2,p<0.001）,but negatively correlated with CIR₃₀ in T2DM group（r=-0.17,p=0.033）,and WHR was always negatively correlated with CIR₃₀ DI（p<0.05）.In multiple stepwise regression analysis,WHR was the independent determinant of CIR₃₀ and CIR₃₀ DI in NGM group and T2DM group(CIR₃₀:β:0.595±0.201,-1.157±0.451;DI:-0.386±0.209,-0.916±0.412).There was significant interaction between WHR and CIR₃₀ under different glucose metabolism（interaction termβ:-0.387±0.275,-1.441±0.313）.（p<0.001）.The results of intermediary analysis showed that BMI,HOMA-IR mediated the compensatory increase of CIR₃₀ caused by BMI in various states of glucose metabolism.For WHR,in the NGM group,HOMA-IR also mediated the compensatory increase in CIR₃₀ caused by WHR（ACME=0.987,p<0.001;mediating effect accounted for 77%）,while in the T2DM group,the effect of WHR on BCF was direct,effect direction was negative（ACME=0.108,p=0.738;ADE=-1.046 BCF 0.004）.There was a high correlation between WHR and the indexes of body fat distribution in reaction center（r=0.35,p<0.05;r=0.35,p<0.05;r=0.37,p<0.05）.Through the analysis of the relationship between body fat distribution index and BCF,it was found that in NGM group,the total fat rate FM and T/E were both positive determinants of CIR₃₀（β:0.012±0.004,0.196±0.098）,while in pre-diabetes/T2DM group,trunk fat was a positive determinant of CIR₃₀,while T/E was a negative determinant of CIR₃₀（β:-0.249±0.127,p<0.05）,while in pre-diabetes/T2DM group,trunk fat was a positive determinant of CIR₃₀,while T/E was a negative determinant of CIR₃₀（β:-0.249±0.127,p=0.031）.Conclusion:Compared with BMI and WC,WHR can more reasonably reflect the changes of obesity-related insulin secretion under different glucose metabolism:in the compensatory increase stage of BCF in NGM,there is a positive correlation between WHR and BCF;in the stage of slow decrease of BCF in prediabetes,there is a negative correlation between WHR and BCF;in the stage of obvious decrease of BCF in T2DM,there is a significant negative correlation between WHR and BCF,but there is no such trend in the correlation between BMI and WC and BCF.And WHR is the only independent obesity index that does not depend on IR and leads to the decrease of BCF.The correlation between WHR and central body fat distribution index is better,which may be the reason why WHR is more able to reflect the BCF of obese patients with different glucose metabolism.Part III Relationship between changes in BCF and anthropometric index after weight lossObjective:To study the changes of BCF in obese patients with different glucose metabolism after weight loss treatment,and the effects of BMI,WC,WHR and other anthropometric indexes on BCF function and glucose metabolism after weight loss.Identify the index with the greatest impact,and try to use this index to build a BCF related prediction model.Methods:Among the 847 obese patients,179 received comprehensive treatment of weight loss and were followed up for 3 to 6months.Finally,79 obese patients were followed up,including 26 patients with NGM,32 patients with prediabetes and 21 patients with T2DM.All subjects underwent a comprehensive assessment of their obesity index,BCF and glucose metabolism before and after weight loss treatment.The changes of obesity index and BCF index after weight loss under different glucose metabolism were analyzed,and the relationship between the relative change value of BCF（ΔBCF）and the baseline value andΔvalue of each obesity index was compared.According to the changes of glucose metabolism after weight loss,some patients were divided into two groups:improved glucose metabolism group and non-improved glucose metabolism group.Through single factor stratified analysis and multiple logistic regression analysis to identify the related obesity index factors that are not conducive to the improvement of glucose metabolism,and establish a prediction model for the improvement of glucose metabolism.Results:After weight loss treatment,CIR₃₀ in diabetic patients increased significantly（p<0.05）.Insulin sensitivity index ISI increased significantly in all patients.The level of DI in patients with prediabetes and T2DM was higher than that before.Correlation analysis showed that there was a negative correlation between baseline WHR andΔCIR₃₀ in NGM group（r=-0.52）,and a negative correlation between BMI,WC,WHR andΔCIR30 in prediabetes group（r=-0.48,r=-0.51,r=-0.36）.The relationship betweenΔBMI,ΔWC,ΔWHR andΔCIR₃₀ andΔCIR₃₀DI is not significant.Stepwise regression analysis showed that baseline WHR level was the only determinant ofΔCIR₃₀（β:-2.136±0.811;p=0.010）.After weight loss treatment,a total of 33 people improved their glucose metabolism,with a total improvement rate of 57.8%.The stratified analysis of single factors affecting the improvement rate of glucose metabolism showed that the higher the baseline WHR,the smaller the glucose metabolism improvement rate（baseline WHR:low,medium,high,corresponding improvement rate 72%,54.5%41%）.Multivariate logistic regression analysis showed that baseline WHR（OR:0.033;95%CI0.011～0.897;p=0.039）and A/G（OR:0.031;95%CI 0.002～0.556;p=0.028）were unfavorable factors affecting the improvement of glucose metabolism.Finally,a prediction model based on age,sex,WHR and glucose metabolism of obese patients was established:ln（P/（1-P））=-2.685*WHR+0.011*age-0.143*sex+0.339*pre DM/0.611*T2DM+3.15.This model has more than 80%accuracy in predicting the improvement of glucose metabolism in obese patients after weight loss.Conclusion:Weight loss in 3 to 6 months improved the BCF of obese patients in varying degrees,mainly reflected in the increase of insulin secretion in T2DM patients,and the improvement of insulin sensitivity and DI in all patients.The higher the baseline WHR of NGM patients,the smaller the improvement of insulin secretion after weight loss.At the same time,high baseline WHR is also an important factor that is difficult to reverse glucose metabolism in pre-diabetes and T2DM patients after weight loss.In the cross-sectional and longitudinal studies of the relationship between BMI and BCF.WHR always shows additional risk increment information that BMI and WC do not have,so WHR is a better indicator of obesity damage to BCF and glucose metabolism.