Quantitative Diagnosis Of Hepatic Fat Content And Serum Cystatin C Level In Different Glucose Metabolism Populations

Background and aims The prevalence of nonalcoholic fatty liver disease（NAFLD）is consistent with a rapid increase in the prevalence of central obesity,type 2 diabetes mellitus（T2DM）,and metabolic syndrome.NAFLD is the clinical manifestation of metabolic syndrome in the liver,and insulin resistance is the most important risk factor.Previous studies have shown that hepatic fat deposition plays a key role in the development of insulin resistance.Studies have confirmed that insulin resistance is present in patients with normal glucose tolerance,and insulin resistance is aggravated with increased hepatic fat content（HFC）.Early intrahepatic fat deposition can also predict the risk of developing type 2 diabetes and its cardiovascular complications in the future.However,there are few reports on the relationship between HFC and glucose metabolism.The current challenge is to determinethe threshold of HFCthat causes abnormal glucose metabolism with hepatic fat deposition,which is the cut-off point of initial intervention and target of treatment.The key to solving this issue is to accurately quantify HFC and explore its quantitative relationship with glucose metabolism.Therefore,it is necessary to carry out quantitative research on the quantitative relationship between HFC and insulin resistance and β cell function damage in different glucose metabolism populations.In this study,oral glucose tolerance test was used to screen out different glucose metabolism populations,and Hydrogen proton magnetic resonance spectroscopy（¹H-MRS）was used to determine HFC in subjects.This study explored the quantitative relationship between quantitative HFC and islet function,and investigated the effect of different degrees of HFC on islet function among different glucose metabolism levels populations,providing a theoretical basis for the prevention and treatment of diabetes with fatty liver.Previous studies have suggested that Cystatin C（Cys C）is an ideal marker for diabetic nephropathy and a risk factor for other chronic complications of diabetes.In recent years,it has been reported that serum Cys C levels increase rapidly in the pre-diabetes stage,and can also predict the risk of diabetes.Animal studies have shown that it has the effect of inducing insulin resistance,and insulin resistance is an important pathogenesis of T2 DM.If there is some intrinsic linksamong hepatic fat deposition,serum Cys C,insulin resistance and diabetes,until now,there is no relevant report on the relationship between quantitative HFC and Cys C and insulin resistance.In the present study,we carried out study on the relationship among HFC,Cys C and insulin resistance in different glucose metabolism populations,and provides theoretical basis for further molecular mechanism research.The aim of this study was to compare the relationship between HFC and serum Cys C in different glucose metabolism populations,and to analyze the relationship between HFC,Cys C and insulin resistance,and to explore the significance of Cys C as a potential serological indicator of HFC,and to provide a theoretical basis for serological indicators in the prevention and treatment of diabetes with fatty liver.NAFLD has become a reserve army for T2 DM and its cardiovascular complications.Early intervention of HFC is of great significance for the prevention and treatment of T2 DM.At present,it is necessary to study the relationship between HFC and glucose metabolism in large-scale population of T2 DM,butthe present measure methods of HFCstillhave some certain defects.Although hydrogen proton magnetic resonance spectroscopy has the advantage of accurate quantification,it has some obvious shortcomings whenhaving large-scale clinical application,such as time-consuming,expensive,and high equipment requirements.Therefore,scholars have actively explored a non-invasive and effective method for quantifying HFC,which has important practical significance for clinical prevention and treatment of T2 DM.In this study,hydrogen proton magnetic resonance spectroscopy（¹H-MRS）was used to measure HFC.It is proposed to explore and establish a new non-invasive,accurate and reliable quantitative ultrasound method for determination of HFC by quantified ultrasound the hepatic/renal echo ratio and hepatic echo attenuation rate,and provide a theoretical basis for the future application of this technology in the prevention and treatment of T2 DM.Methods Study 1: 1.The 141 patients with newly diagnosed T2DM（NT2DM）,48 pre-diabetes subjects（PDM）and 53 normal control（NC）with normal glucose tolerance were recruited from Hefei Hospital Affiliated to Anhui Medical University（the Second People’s Hospital of Hefei）,including 125 males and 117 females.2.All subjects had a questionnaire survey and measured for some parameters,including height,weight,waist circumference,hipline,systolic blood pressure,diastolic blood pressure,BMI and WHR.Venous blood was collected for more than 10 hours overnight,The subjects were tested for the following items: fasting plasma glucose（FPG）,total cholesterol（TC）,TG,high-density lipoprotein-cholesterol（HDL-C）,low-density lipoprotein-cholesterol（LDL-C）,very low-density lipoprotein-cholesterol（VLDL-C）,alanine aminotransferase（ALT）,alkaline phosphatase（ALP）,aspartate aminotransferase（AST）,γ-glutamyltransferase（GGT）,lactate dehydrogenase（LDH）,total bilirubin（TBIL）,indirect bilirubin（IBIL）,direct bilirubin（DBIL）,creatinine,uric acid（UA）,apolipoprotein（Apo）-A1 and Apo-B.All the blood biochemical indexes were determined 3.The OGTT was performed after a 10 hours fast with venous blood sampling in the fasting state,then,fasting plasma glucose（FPG）,30 min plasma glucose,60 min plasma glucose,120min（2h）plasma glucose,fasting insulin,30 min insulin,60 min insulin,120 min insulin,fasting C-peptide,30 min insulin C-peptide,60 min C-peptide and 120 min insulin C-peptide were measured.Blood glucose was detected by hexokinase method and insulin and C-peptide were used by chemiluminescence method,and blood glucose was determined to be enrolled according to the OGTT.It was assessed insulin resistance and islet beta cell function by homeostasis model assessment ofinsulin resistance（HOMA-IR）and homeostasis model assessment ofbeta cell（HOMR-β）and insulin sensitivity index（ISImatsuda）.The difference value between 30 min plasma glucose,30 min insulin,30 min C peptide and fasting in OGTT was Delta G30,Delta Ins30 and Delta CP30.Early phase function of islet β cell: the difference between 30 min insulin and its fasting in OGTT/the difference between 30 min plasma glucose and fasting（Ins30/G30）and the difference between 30 min C peptide in OGTT and its fasting / 30 min plasma glucose and its fasting Value（CP30/G30）.Whole function of islet-βcell: area under curve of insulin in OGTT /area under curve of blood glucose(Ins _AUC/G_AUC)and area under curve of C peptide in OGTT / area under curve of blood glucose(CP_AUC /G_AUC).4.HFC was measured by ¹H-MRS.All subjects were divided into four subgroups of Q1,Q2,Q3 and Q4 according to HFC quartile distribution.5.Data collecting and logging was carried out using electronic spreadsheet code Excel,and statistical analysis was performed in using SPSS（Statistical Package for the Social Sciences）version 17.0.6.To explore the quantitative relationship between quantitative HFC and insulin resistance and islet β cell function.Study 2: 1.The selection and group of study subjects were described in the first part of the paper.All subjects were surveyed and measured for parameters and detected for blood biochemical and hepatic fat content by ¹H-MRS in the first part of the paper.2.Serum Cys C levelwas measured by immunoturbidimetric method.3.All subjects were divided into four subgroups of Q1,Q2,Q3 and Q4 according to HFC quartile distribution.4.Data collecting and logging was processed by using electronic spreadsheet code Excel,and statistical analysis was carried out using SPSS version 17.0.5.Study 3: 1.The selection and group of study subjects were described in the first part of the paper.All subjects were surveyed and measured for parameters and detected for blood biochemical and HFC by ¹H-MRSin the first part of the paper.2.When divided HFC with per 5% increasing,the HFC of the entire study subject could be allocated into five groups,HFC <5%,5% ≤ HFC <10%,10% ≤ HFC <15%,15% ≤ HFC<20 % and 20% ≤ HFC according to the 5% of hepatic fat content.3.All subjects were detected by abdominal ultrasound examination.The image analysis software was used to determine the hepatic/renal echo ratio and hepatic echo attenuation rate on the ultrasound image.The simulated natural human 3D abdominal model was used to adjust ultrasound parameters to obtain the quantified ultrasound hepatic/renalratio（QUS HRR）and quantified ultrasound hepatic echo-intensity attenuation rate（QUS HAR）.4.Data collecting and logging was carried out using electronic spreadsheet code Excel,and statistical analysis was carried out using SPSS version 17.0 and Med Calcversion 11.4.2.0.5.It was investigated the relationship between HFC by ¹H-MRS and quantitative ultrasound parameters,and established a quantitative prediction model for HFC by quantitative ultrasound parameters,body parameters and cystatin c.It was analyzed by Bland-Altman analysis between HFC by ¹H-MRS and HFC by quantitative ultrasound.It was determined the ROC analysis of NAFLD by quantitative ultrasound method.The association of HFC by ¹H-MRS and HFC by quantitative ultrasound in obese subgroup with BMI≥28.0 was compared.The correlation between HFC by quantitative ultrasound and clinical parameters was established.Intraclass correlation coefficient and Bland-Altman analysis were used for quantitative ultrasound parameters among different operators and different ultrasound instruments.Results: Study 1 1.The distribution of HFC concentration among NC,PDM and NT2 DM groups was 5.11% ± 3.66%,9.62% ± 4.41%,14.72% ± 6.37%,respectively.There were significant differences of BMI,WHR,SBP,DBP,UA,IBIL,ALT,GGT,LDH,TC,TG,HDL-C,LDL-C,VLDL-C,Apo-A1,FPG,2h PG,HOMA-IR,HOMA-β and HFC（all P<0.05）.In addition,the level of HFC and HOMA-IR was orderly increasing in NC,PDM and NT2 DM groups（all P<0.0125）,however,there was an orderly decreased HOMA-β level in NC,PDM and NT2 DM groups（all P<0.0125）.2.Correlation analysis indicated that,in NT2 DM group,HFC was positively associated with BMI,SBP,DBP,FPG and HOMA-IR.,negatively associated with TBIL,DBIL and IBIL（all P<0.05）.In PDM group,HFC had an association with BMI,FPG,UA and HOMA-IR（all P<0.05）.Moreover,there was a positive association of HFC with FPG,BUN,ALT,ALP and HOMA-IR,negative association of HFC with Apo-A1（all P<0.05）.3.Based on the distribution of HFC,we calculated the quartile of LFC and divided it into four groups（Q1: LFC < 5.89%,Q2: 5.89% ≤ LFC < 11.62%,Q3: 11.62% ≤ LFC < 16.26% and Q4: 16.26% ≤ LFC）.The results indicated that there was a significant difference of FPG,2h PG,Ins30/G30,CP30/G30,Ins _AUC/G _AUC,CP _AUC /G _AUC,HOMA-IR,HOMA-β and ISImatsuda among four quartile groups（all P < 0.05）.The FPG,2h PG and HOMA-IR was increased from Q2,and reached to the peak in Q4,while,Ins30/G30,CP30/G30,Ins _AUC/G _AUC and HOMA-β was gradually decreased from Q2,and had its lowest value in Q4.4.The results of correlation analysis showed that HFC had positive relations with FPG,2h PG,HOMA-IR and Delta G30（all P < 0.001）,but had negative associations with HOMA-β、ISImatsuda、Ins30/G30、CP30/G30、Ins _AUC/G_AUC、CP_AUC /G_AUC、Delta Ins30、Delta CP30（all P < 0.01）5.Mutilinear regression with step-wise method suggested that HFC is one of the influential factors associated with HOMA-IR.Study 2 1.There was a significant difference of Cys C in NC,PDM and NT2 DM groups（all P < 0.05）,in compared with NC group,both the PDM and NT2 DM groups had an elevated Cys C level（all P < 0.001）.2.According to the distribution of HFC,we calculated the quartile of LFC and divided it into four groups（Q1: LFC < 5.89%,Q2: 5.89% ≤ LFC < 11.62%,Q3: 11.62% ≤ LFC < 16.26% and Q4: 16.26% ≤ LFC）.When comparing the difference of Cys C among four groups,we found that there was statistical significantly difference of Cys C（all P<0.05）.In compared to Q1 group,the level of Cys C was significantly increased in both Q3 and Q4 group（all P<0.0083）.However,we did not observed any Cys C difference among other comparing groups.3.Correlation analysis revealed that serum Cys C concentration was significantly associated with BMI,WHR,SBP,DBP,2h PG,Cr,IBIL,ALT,GGT,ALP,TG,TC,LDL-C,HDL-C,VLDL-C,Apo-A1,HOMA-IR and HFC（all P<0.05）.In addition,we also analyze the relation of HFC and NOMA-IR with clinical and laboratory parameters,the results showed that HFC correlated with age,BMI,WHR,SBP,DBP,FPG,2h PG,UA,ALT,GGT,TG,TC,HDL-C,LDL-C,VLDL-C and Apo-A1（all P<0.05）,HOMA-IR correlated with BMI,WHR,SBP,DBP,FPG,2h PG,TG,TC,HDL-C,LDL-C and VLDL-C（all P<0.05）.4.Mutilinear regression indicated that WHR,FPG,TC and Cys C are the predictors for HFC,BMI,DBP,FPG,Cys C and HFC are the predictors for the HOMA-IR.Study 3 1.When divided HFC with per 5% increasing,the HFC of all the study subject could be allocated into five groups,which included HFC <5%（n=48）,HFC:2.96%±1.12%;5% ≤ HFC <10%（n=65）,HFC:7.39%±1.33%;10% ≤ HFC <15%（n=50）,HFC:13.18%±1.42%;15% ≤ HFC<20%（n=55）,HFC:17.03%±1.33%;20% ≤ HFC（n=24）,HFC:24.60%±2.86%.There was significant difference of height,body mass,BMI,waistline,hipline,WHR,SBP,DBP,FPG,2h PG,BUN,UA,TBIL,IBIL,ALT,GGT,TG,TC,HDL-C,LDL-C,VLDL-C,Apo-A1,HFC,QUS HRR and QUS HAR（all P<0.05）.2.Linear correlation analysis found that QUS HRR and QUS HAR were positively associated with HFC quantified by ¹H-MRS,respectively（r=0.946,P<0.001 and r=0.936,P<0.001）.3.The results of Mutilinear regression discovered that the application of QUS HRR was able to independently estimate the HFC with adjusted R2 for 0.895（P < 0.001）.When combined QUS HRR with QUS HAR,the total model adjusted R2 for 0.935（P < 0.001）.The prediction formula: HFC（%）= 28.956 × QUS HRR + 218.045 × QUS HAR – 8.892.When combined Cystain C with the prediction model,the total adjusted R2 for 0.936（P < 0.001）.When combined age,gender,BMI and WHR parameters with the prediction model,the total adjusted R2 for 0.937（P < 0.001）.When combined age,gender,BMI,WHR and Cystatin C with the prediction model,the total adjusted R2 for 0.937（P < 0.001）.4.For the entire study subject,both the ¹H-MRS and Quantified Ultrasound were applied for HFC detection.HFC baseline was defined by ¹H-MRS（95% limits of agreement:-3.3790,3.3777）,Bland-Altman analysis was performed in HFC <11.12% subgroup population,the results observed that Quantified Ultrasound overestimated 0.54% the baseline of HFC in compared with ¹H-MRS,with 95% limits of agreement from-3.6761 to 2.5857.5.In compared with 5.56% of HFC quantified by ¹H-MRS as NAFLD diagnosis standard,the Quantified Ultrasound found that HFC of the diagnosis standard for NAFLD is 6.71% with a sensitivity 94.15% and specificity 96.3% by ROC analysis.When analyzing the ¹H-MRS quantified HFC <11.12% study population,the results of ROC analysis indicated that the sensitivity and specificity were 95.31% and 90.74% when applied Quantified Ultrasound in the diagnosis of NAFLD.6.We stratified the study population according to BMI ≥28 as subgroup,the results demonstrated that there was a strong relationship between ¹H-MRS and Quantified Ultrasound in determination of HFC.7.Partial correlation analysis was determined by adjusting the influence of age,BMI and gender ratio,the results observed that HFC by Quantified Ultrasound was positively correlated with WHR,SBP,FPG,ALT,UA,TC,LDL-C,and negatively correlated with LDH（all P<0.05）.8.Intraclass correlation coefficient between QUS HRR and QUS HAR among different operators and ultrasonic instruments was 0.970,0.978（all P<0.001）and 0.972,0.981（all P<0.001）.When analyzing the repeatability among different operators and ultrasonic instruments in quantified HFC,the results of Bland-Altman analysis indicated thatthe difference value was 97.0% and 95% within ±2.5% error rangein different operators and ultrasonic instruments with quantified HFC.Conclusions: Study 1: When the HFC was 5.89%,pancreatic β cell function was slightly impaired with the emerge of insulin resistance and small increase of BG.Once the HFC reached to 11.62%,insulin resistance apparently aggravated,pancreatic β cell function was seriously impaired with an elevated BG concentration,and met the diagnosis standard criteria for DM.Our finding suggested that the deposition of HFC may involve in the pathogenesis and development of T2 DM through responding strong insulin resistance.Study 2: There was an increase of Cys C level in the stage of PDM and HFC >11.62%,suggesting that Cys C may be an indicator for the prediction of HFC and HOMA-IR.Study 3: In compared to ¹H-MRS,the use of Quantified Ultrasound for quantification of HFC is more practical,accurate and reliable.Quantified Ultrasound is capable to provide a potential substitution for HFC detection,and make it possible to quantified HFC in a larger T2 DM population.The role of serum cystatin C in predicting HFC is limited.