Relationship between liver fat content and liver enzymes in individuals with various statuses of glucose metabolism

：Objective To study therelationship between liver fat content (LFC) and liver enzymes in individuals with variousstatuses of glucose metabolism. Methods A total of 109 subjects including with impairedglucose regulation (IGR) ( n = 31 ), newly diagnosed type 2 diabetes (NT2DM) ( n = 31 )and normal glucose tolerance (NGT) (n =47) were recruited. The level of LFC was measuredby 1H magnetic resonance spectroscopy (1H-MRS) to study the relationship between liver fatcontent (LFC) and alanine aminotransferase ( ALT ), aspartate aminotransferase ( AST ),alkaline phosphatase ( AKP ) and γ-glutamyltransferase(GGT).The receiver operating characteristic curve(ROC) was employed to obtain the optimalcut-off point of ALT to predict the occurrence of nonalcoholic fatty liver disease(NAFLD). Results (1)The levels of LFC were progressively raised in NGT, IGR and NT2DMgroups respectively [3.83(2.35-7.59)%, 12.82 (8.10-21.37)% and 21.99 (11.89-34.43)%, P＜0.01];(2) the subjects were divided into four subgroups by the method of LFC quartile. Andquartile subgroups Q1 -4 were associated with the increase of LFC. Waist, BMI, systolicblood pressure, triglyceride, total cholesterol,fasting plasma glucose, OGTT 2 hpostprandial glucose and HOMA-IR had a rising trend from Q2. But HDL-C showed a decliningtrend from Q2; (3) ALT and GGT singnificantly increased from Q3( P ＜0.01 )while AST and AKP significantly increased in Q4( P ＜0. 01 ); (4) adjusted bygender, age and body mass index ( BMI), LFC was positively correlated with AST ( r = 0.329, P ＜ 0. 05 ), ALT ( r = 0. 454)and GGT ( r = 0. 378 ) ( All P ＜ 0. 01 ). But it wasnegatively correlated with AST/ALT ( r = - 0. 364, P ＜ 0. 01 ); (5)the analysis of stepwise regression demonstrated that LFC was apredictor of ALT, AST, GGT and AST/ALT; (6) ALT had a ROCAuc of 0. 813 (male) and 0. 769(female) ( All P ＜ 0. 01 ). The optimal cut-off point of 23. 5 U/L(male) and 17. 5 U/L(female) mightpredict the occurrence of NAFLD. Conclusions Liver enzymes are correlated with LFC even innormal range. The optimal cut-off point of 23. 5 U/L(male)and 17. 5 U/L (female) mightpredict the occurrence of NAFLD. The current used ALT upper limit could underestimate theNAFLD. Key words: Glucose metabolism disorders; Fatty liver; Diabetesmellitus; Magnetic resonance spectroscopy; Liver function tests