Retrospective Analysis Of The Prevalence Of Fatty Liver Disease In Hospitalized Diabetic Patients

Objective:To explore the detection rate of non-alcoholic fatty liver diseases (NAFLD) in patients with type2diabetes mellitus(T2DM) applying the diagnostic criteria of the combination of alanine aminotransferase (ALT)、aspartate aminotransferase (AST) and γ-glutamyltranspeptidase (GGT), and factors influencing in T2DM associated with NAFLD.Methods:Screen the history information of total1234hospitalized patients with diabetes mellitus (DM) in the department of endocrinology of Zhongshan hospital Fudan University from September2010to October2011. Collect demographic data(include gender, age, DM duration, alcoholic history and so on), clinical data (include height, weight, waist circumference, hip circumference, systolic blood pressure, diastolic blood pressure and so on), laboratory examination[include carbohydrate metabolism (such as fasting blood glucose, hemoglobin Ale, glycated albumin, fasting insulin and fasting C peptide), lipid metabolism (such as total cholesterol, triglyceride, high-density lipoprotein, low-density lipoprotein, non-high-density lipoprotein, apolipoprotein A, apolipoprotein B, apolipoprotein E and lipoprotein a), liver function (such as alanine aminotransferase, aspartate aminotransferase and γ-glutamyltranspeptidase), uric acids, C-reactive protein and24-hour urinary protein excretion. Calculate the body mass index (BMI), homeostasis model assessement-insulin resistance (HOMA-IR). Define elevated liver enzymes as any elevation in ALT, AST or GGT above the laboratory cutoffs used in Zhongshan hospital, that is, an ALT≥75U/L, or AST≥75U/L, or GGT>50U/L Exclude from the entire study any one with history of viral hepatitis or positive hepatitis B antigen or hepatitis C antibody, average intake of more than140g alcohol in men or70g alcohol in women per week, application of liver injury drugs, autoimmune hepatitis, any endocrine disease may induce fatty liver such as hypothyroidism and malignant tumor, and the rest with elevated liver enzymes were considered as NAFLD. We analysed:1the detection rate of elevated liver enzymes in patients with diabetes;2the detection rate of elevated liver enzymes in patients with T2DM;3the detection rate of NAFLD in patients with T2DM and excluding any other diseases inducing fatty liver.Results1. The proportion of patients with T1DM, T2DM, MODY and gestational diabetes and any other special types of diabetes were7.4%,90.4%,0.1%and2.1%respectively.2. The detection rate of elevated liver enzymes in patients with diabetes, T1DM, T2DM and gestational diabetes and any other special types of diabetes were28.3%,12.6%,28.7%and60.0%.3. There were31patients with viral hepatitis (9.3%),31with alcoholic hepatitis(9.3%),20with suspected alcoholic hepatitis(6.0%),3with usage of liver injury hi story (0.9%),1with history of schistosomiasis (0.3%),11with hyperthyroidism(3.3%),9with hypothyroidism(2.7%),5with Cushing syndrome(1.5%) and24with malignant tumor(7.2%) in334patients with the combination of diabetes and elevated liver enzymes. The rest199patients were considered to be with NAFLD with the proportion of59.6%.4. There were31patients with viral hepatitis（10.1%),29with alcoholic hepatitis(9.4%),18with suspected alcoholic hepatitis(5.9%),2with usage of liver injury history(0.7%),11with hyperthyroidism(3.6%),10with hypothyroidism(3.3%) and21with malignant tumor(6.8%) in307patients with the combination of T2DM and elevated liver enzymes. The rest185patients were considered to be with NAFLD with the proportion of59.6%.5. There were185cases of NAFLD with detection rate of24.4%in patients with T2DM. The detection rate of NAFLD does not have significant difference between different genders. Both ALT (γs=-0.130) and GGT (γs=-0.125) were negatively correlated with gender(P<0.01in both). There were no correlation between AST, NAFLD fibrosis score and gender.6. ALT (γs=-0.130) and GGT (γs=-0.125) were correlated with gender (men), AST had no correlation with gender, and the detection rate had no difference in different genders.7. With the increasing of age, ALT, AST and GGT reduced significantly, and the detection rate of NAFLD in total population and in men both reduced significantly(trend χ2=17.500,13.556; P<0.01in both), and no similar trend in women. ALT(γs=-0.255), GGT (γs=-0.127)(P<0.01in both) and AST (γ s=-0.084, P<0.05) were all negatively correlated with age.8. The detection rate of NAFLD in new diagnostic diabetic patients and diagnosed diabetic patients were32.4%and24.0%which were not significantly different. ALT(γs=-0.272), AST (γs=-0.162) and GGT (γs=-0.278) were negatively correlated with diabetes duration.9. Patients with NAFLD had higher hemoglobin (P<0.01), WBC(P<0.05) and CRP (P<0.01) than patients without NAFLD.10. ALT was positively correlated with FINS (γs=0.204), FCP(γs=0.311), HOMA-IR (γs=0.254)(P<0.01in all); AST was positively correlated with FINS (γ=0.176), FCP (γ=0.261),HOMA-IR(γs=0.176)(P<0.01in all) and negatively correlated with GA(γs=-0.094, P<0.01);GGT was positively correlated with FBG (γs=0.107), FINS (γs=0.200), FCP (γs=0.381), HOMA-IR (γs=0.352).11.Set whether NAFLD or not as dependent variation, the age, gender, duration of diabetes, BMI, waist circumference, hip circumference, DBP, FBG, HbAlc、GA、FINS、FCP、HOMA-IR、TC、TG、HDL-ch、non-HDL、APO-A、APO-B、APO-E、 LP (a)、CRP and UA as independent variation, BMI and DBP entered in the Logistic stepwise regression analysis at last, and the OR(95%CI) were1.202(1.026～1.407) and1.054(1.001～1.109) respectively.Conclusion1. The detection rate of elevated liver enzymes in DM, T1DM and T2DM were28.3%,12.6%and28.7%respectively.2. The proportion of NAFLD both in DM patients and in T2DM patients were more than the half of the total number of elevated liver, suggesting NAFLD was the primary cause of elevated liver enzymes in DM and T2DM patients.3. The detection rate of NAFLD in hospitalized patients with T2DM applying the combination of ALT, AST and GGT was almost a quarter, which remind us notice the early screening of liver enzymes in patients with T2DM.4. There were negative correlations between age and duration with NAFLD, which remind us notice the early multi-factor intervention in patients with T2DM.5. The independent risk factors of NAFLD in patients with T2DM were BMI and DBP. Objective:To explore the detection rate of non-alcoholic fatty liver diseases (NAFLD) in patients with type2diabetes mellitus(T2DM) applying the diagnostic criteria of hepatic ultrasonography, and factors influencing in T2DM associated with NAFLD.Methods:Screen the history information of total1234hospitalized patients with diabetes mellitus (DM) in the department of endocrinology of Zhongshan hospital Fudan University from September2010to October2011, and focus on the patients who had taken hepatic ultrsound. Collect demographic data, clinical data, laboratory examination; hepatic ultrasonography. Calculate the body mass index (BMI), homeostasis model assessement-insulin resistance(HOMA-IR). Exclude from the entire study any one with history of viral hepatitis or positive hepatitis B antigen or hepatitis C antibody, average intake of more than140g alcohol in men or70g alcohol in women per week, application of liver injury drugs, autoimmune hepatitis, any endocrine disease may induce fatty liver such as hypothyroidism and malignant tumor, and the rest with fatty liver diagnosed by liver ultrasound were considered as NAFLD. We analysed:1the detection rate of fatty liver in patients with diabetes;2the detection rate of fatty liver in patients with T2DM;3the detection rate of NAFLD in patients with T2DM and excluding any other diseases inducing fatty liver.Results1. The proportion of patients with T1DM, T2DM, MODY and gestational diabetes and any other special types of diabetes were6.8%,91.3%,0.1%and1.8%respectively.2. The detection rate of fatty liver in patients with diabetes, T1DM, T2DM and gestational diabetes and any other special types of diabetes were55.0%,33.3%,56.7%and47.1%. 3. There were40patients with viral hepatitis(7.5%),28with alcoholic hepatitis(5.3%),36with suspected alcoholic hepatitis(6.8%),1with usage of liver injury history(0.2%),9with hyperthyroidisra(1.7%),20with hypothyroidism(3.8%),3with Cushing syndrome(0.6%) and21with malignant tumor(4.0%) in531patients with the combination of diabetes and fatty liver diagnoses by hepatic ultrasound. The rest373patients were considered to be with NAFLD with the proportion of70.2%.4. There were39patients with viral hepatitis(7.8%),28with alcoholic hepatitis(5.6%),36with suspected alcoholic hepatitis(7.2%),1with usage of liver injury drug history (0.7%),9with hyperthyroidism(1.8%),1with hypothyroidism(0.2%),1with Cushing syndrome(0.2%) and20with malignant tumor(4.0%) in500patients with the combination of T2DM and fatty liver diagnosed by hepatic ultrasound. The rest347patients were considered to be with NAFLD with the proportion of72.0%.5. There were356cases of NAFLD with the detection rate of56.8%in patients with T2DM; the detection rate of NAFLD did not exist a significant difference between different genders; no statistical association between gender and NAFLD.6. With the increasing of age, and the detection rate of NAFLD in total population and in men both reduced significantly(trend χ2=10.604,9.715; P <0.01in both), and no similar trend in women.7. The detection rate of NAFLD in new diagnostic diabetic patients and diagnosed diabetic patients were68.2%and56.4%respectively. The detection rate of NAFLD does not exist significant difference between the two groups. The statistical association exist between diabetic duration and NAFDL[OR(95%CI)0.965(0.944-0.986), P<0.01]8. Patients with NAFLD had higher hemoglobin, RBC(P<0.01in all) and WBC(P <0.05) than patients without NAFLD.9. After adjusting for gender, age and duration of diabetes, the waist circumference (P=0.012)、WHP (P=0.018)、HOMA-IR (P=0.000)、FCP (P=0.000)、 TG(P=0.007)、non-HDL(P=0.014)、APO-B（P=0.002)、APO-E(P=0.023)、ALT (P=0.000)、 AST(P=0.003)、GGT(P=0.000)、UA(P=0.000)and average of24h uric albumin(P=0.042) in patients with NAFLD were still higher than those in patients without NAFLD, and the difference were significant. 10. Set whether NAFLD or not as dependent variation, BMI, TG and HDL-ch entre in the Logistic stepwise regression analysis at last, and the OR(95%CI) were1.342(1.173～1.535)、1.207(1.03～1.412)�'�0.126(0.025～0.622)respectively.Conclusion1. The detection rate of fatty liver in DM, T1DM and T2DM were55.0%,3.3%and56.7%respectively.2. The proportion of NAFLD both in DM patients and in T2DM patients were more than70%of the total number of fatty liver, suggesting NAFLD was the primary cause of fatty liver in DM and T2DM patients.3. The detection rate of NAFLD in hospitalized patients with T2DM applying liver ultrasonography was more than a half, which remind us pay attention to the early screening and treatment of NAFLD in patients with T2DM.4. There were negative associations between age and duration with NAFLD, which remind us notice the early multi-factor intervention in patients with T2DM.5. The risk factors of NAFLD in patients with T2DM were:BMI and TG, and the protective factor was HDL-ch. Objective:To explore the distribution of liver fatty content and detection rate of non-alcoholic fatty liver diseases (NAFLD) in patients with hospitalized type2diabetes mellitus(T2DM) applying the diagnostic criteria of the, and factors influencing in T2DM associated with LFC.Methods:Screen the history information of total1234hospitalized patients with diabetes mellitus (DM) in the department of endocrinology of Zhongshan hospital Fudan University from September2010to October2011, and focus on the patients who had taken hepatic’H magnetic resonance spectroscopy. Collect demographic data, clinical data, laboratory examination;1H magnetic resonance spectroscopy of liver. Calculate the body mass index (BMI),homeostasis model assessement-insulin resistance(HOMA-IR). LFC=area of fat peak*100/(area of fat peak+area of water peak). Exclude from the entire study any one with history of viral hepatitis or positive hepatitis B antigen or hepatitis C antibody, average intake of more than140g alcohol in men or70g alcohol in women per week, application of liver injury drugs, autoimmune hepatitis, any endocrine disease may induce fatty liver such as hypothyroidism and malignant tumor, and the rest with fatty liver diagnosed by liver ultrasound were considered as NAFLD. We analysed:1the detection rate of fatty liver in patients with diabetes;2the detection rate of fatty liver in patients with T2DM;3the detection rate of NAFLD in patients with T2DM and excluding any other diseases inducing fatty liver.Results1. The number of cases of patients with T1DM, T2DM and gestational diabetes and any other special types of diabetes were16,202and2, with the proportion of7.6%,95.7%and1.4%respectively.2. The number of cases with fatty liver in patients with diabetes, T1DM, T2DM and gestational diabetes and any other special types of diabetes were531,10,227and4, and the detection rate of fatty liver were55.0%,41.7%,72.8%and100%.3. There were22patients with viral hepatitis(9.1%),20with alcoholic hepatitis(8.3%),15with suspected alcoholic hepatitis(6.2%),1with usage of liver injury history(0.4%),9with hypothyroidism(3.3%),1with Cushing syndrome(0.4%) and5with malignant tumor(2.1%) in241patients with the combination of diabetes and fatty liver diagnoses by hepatic1H MRS. The rest196patients were considered to be with NAFLD with the proportion of70.1%.4. There were22patients with viral hepatitis(9.7%),18with alcoholic hepatitis(7.9%),15with suspected alcoholic hepatitis(6.6%),1with usage of liver injury drug history (0.4%),7with hypothyroidism(3.1%) and5with malignant tumor(2.2%) in227patients with the combination of T2DM and fatty liver diagnosed by hepatic1H MRS. The rest159patients were considered to be with NAFLD with the proportion of70.0%.5. There were164cases of NAFLD with the detection rate of70.4%in patients with T2DM; the detection rate of NAFLD did not exist a significant difference between different genders; no statistical association between gender and LFC.6. With the increasing of age, the LFC reduce significantly. LFC was negatively correlated with age and after adjusting the confounding factors the correlation still exist.7. The detection rate of NAFLD in new diagnostic diabetic patients and diagnosed diabetic patients were100%and68.8%respectively. The former higher than the later, and the difference was significant. The statistical association exist between diabetic duration and LFC after adjusting confounding factors.8. After adjusting for gender, age, duration of diabetes and BMI, DBP (P=0.015), FCP (P=0.031), TG (P=0.003)、ALT (P=0.010)、AST (P=0.030) and GGT (P=0.000) in patients with NAFLD were still higher than those in patients without NAFLD, and the HDL-ch(P=0.012*) and NAFLD fibrosis score (P=0.007) were lower than those in patients without NAFLD, and the difference were significant.9. LFC was positively correlated with BMI (γs=0.354), waist circumference (γs=0.256), hip circumference(γs=0.235), FINS(γs=0.244), FCP (γs=0.339), HOMA-IR (γs=0.287), TC (γs=0.212), TG (γs=0.443), non-HDL (γs=0.284), APO-B (γs=0.238), APO-E (γs=0.406), UA (γs=0.255), ALT (γs=0.537), AST (γ s=0.459),GGT(γs=0.481)(P<0.01in all) and DBP (γs=0.164, P<0.05); and negatively correlated with LP (a)(γs=-0.300), HDL-ch (γs=-0.165)(P<0.01in both) and HbAlc (γs=-0.287)10. Set LFC as dependent variation, APO-B, ALT and the age entre in the liner stepwise regression analysis at last, and the β were0.352,0.740and0.305respectively.Conclusion1. The detection rate of fatty liver in DM, T1DM and T2DM were70.9%,41.7%and72.8%respectively.2. The proportion of NAFLD both in DM patients and in T2DM patients were more than70%of the total number of fatty liver, suggesting NAFLD was the primary cause of fatty liver in DM and T2DM patients.3. The detection rate of NAFLD in hospitalized patients with T2DM applying1H MRS was70.4%, which remind us pay attention to the early screening and treatment of NAFLD in patients with T2DM.4. There were negative associations between age and duration with NAFLD, which remind us notice the early multi-factor intervention in patients with T2DM.5. The independent risk factors of LFC in patients with T2DM were:APO-B and ALT. ObjectiveTo compare the detection rate of NAFLD in patients with diatetes mellitus using elevated liver enzymes levels, hepatic ultrasonography and proton magnetic resonance spectroscopy (1H MRS).MethodsCollect the history informations of total805hospitalized patients with diabetes mellitus in the department of endocrinology of Zhongshan hospital Fudan University from September2010to October2011. Exclude the ones with history of viral hepatitis or positive hepatitis B antigen or hepatitis C antibody, average intake of more than140g alcohol in men or70g alcohol in women per week, application of liver injury drugs, autoimmune hepatitis, any endocrine disease may induce fatty liver such as hypothyroidism and malignant tumor. Collect the liver function (such as alanine aminotransferase, aspartate aminotransferase and γ-glutamyltranspeptidase), result of hepatic ultrasound and liver fatty content (LFC) by1H MRS. Define elevated liver enzymes as any elevation in ALT, AST or GGT above the laboratory cutoffs used in Zhongshan hospital, that is, an ALT≥75U/L, or AST≥75U/L, or GGT>50U/L. Define LFC≥5.5%as fatty liver using1H MRS.Results1. The detection rate of NAFLD using the combination of elevated liver enzymes, hepatic ultrasonography and hepatic1H MRS were23.1%,54.9%and69.7%. The detection rate of NAFLD with1H MRS was higher than that with hepatic ultrasound, and the later was higher than that with the combination of the elevated liver enzymes.2. Set the hepatic ultrasound as reference standard of NAFLD diagnosis, the sensitivity and specificity of the combination of elevated enzymes were33.8%and81.0%respectively.3. Set the hepatic1H MRS as the non-invasive gold standard of NAFLD diagnosis, the sensitivity and specificity of the combination of elevated enzymes were33.6%and83.8%respectively, and those of hepatic ultrasonography were73.9%and72.1%.4. Set the hepatic ultrasound as reference standard of NAFLD diagnosis, ALT had a ROCAEC of0.610(male) and0.734(female)(All P<0.05). The optional cut-off points of ALT were22.95U/L(male) and19.05U/L(female). GGT had a ROCAUC of0.668(male) and0.737(female)(All P<0.05), and the optional cut-off point of GGT were26.9U/L and28.65U/L.5. ALT had a ROCAUC of0.689(male) and0.727(female)(All P<0.01). The optional cut-off points of ALT were19.65U/L(male) and16.95U/L (female). GGT had a ROCNUC of0.691(male) and0.761(female)(All P<0.01). The optional cut-off points of GGT were27.15U/L(male) and25.9U/L (female). AST had a ROCAUC of0.654and optional cut-off points of16.5U/L.Conclusion1. The detection rate of NAFLD using the combination of elevated liver enzymes, hepatic ultrasonography and hepatic1H MRS were23.1%,54.9%and69.7%. The detection rate of NAFLD with’H MRS was higher than that with hepatic ultrasound, and the later was higher than that with the combination of the elevated liver enzymes.2. Set the hepatic ultrasound as reference standard, the optional cut-off point of ALT to diagnosis NAFLD in DM was22.95U/L in men and19.05U/L in women; and those of GGT were28.65U/L in men and26.9U/L in women.3. The existing cut-off point underestimates the real prevalence of NAFLD. The optinal cut-off point of ALT to diagnosis NAFLD in DM was19.65U/L in men and16.95U/L in women; and those of GGT were27.15U/L in men and25.9U/L in women, and that of AST was16.5U/L.4. The sensitivity and specificity of hepatic ultrasound was73.9%and72.1%respectively. The hepatic ultrasound is still more truly reflect the prevalence of NAFLD.