| Objective: To evaluate the effects of different measurement methods for human advanced glycation endproducts(AGEs)exposure levels,and glucose and lipid metabolism indices were applied to further clarify the associations for their application in large-scale epidemiological studies.Methods: The subjects were residents who underwent physical examination at Shenzhen community health service center in 2019.Demographic characteristics,physical examination characteristics and biological samples were collected.Liquid chromatography tandem mass spectrometry(LC-MS/MS)was used to measure free carboxymethyl lysine(CML),carboxyethyl lysine(CEL),methylglyoxalhydroimidazolone-1(MG-H1),protein-bound CML and CEL.Meanwhile,enzyme linked immunosorbent assay(ELISA)was adopted for total AGEs and AGE Reader was applied for skin autofluorescence(SAF).The basic characteristics of the subjects were statistically described by t-test and chi-square test.Spearman correlation test was introduced to explore the correlation between different determination methods.After adjusting for possible confounding factors,multivariate linear regression analysis was employed to describe the correlation between AGEs and glucose and lipid metabolism indices.Furthermore,the correlation trend was visually displayed by a smooth curve fitting with a generalized additive model.Results:1.A total of 341 subjects,with an average age of 52.27 years,male to female ratio= 1:1.69,were enrolled in this study.The levels of plasma free CML,CEL,MG-H1 and protein-bound CML and CEL measured by LC-MS/MS were 109.48(68.61-148.15),16.05(12.67-18.55),175.50(98.90-226.75),1621.49(1275.00-1871.00),97.91(69.20-110.80)μg/L,respectively.The plasma total AGEs level determined by ELISA was 13.23(3.60-18.76)AU,and SAF level determined by AGE Reader is 1.94(1.67-2.12)AU.2.Spearman correlation analysis showed that there was no correlation between total AGEs with SAF or LC-MS/MS results(P >0.05),but when the measurement value was greater than 5 AU,the total AGEs and free AGEs were weakly correlated(0.1 <r<0.3,P <0.05).A weak correlation between SAF and free AGEs(0.1 <r <0.3,P <0.05)was observed,but there was no correlation between SAF and ELISA(P >0.05).3.Multivariate linear regression analysis showed that free CML was associated with insulin(β =-0.95,P <0.05),LDL-C(β = 0.11,P <0.05),HDL-C(β = 0.04,P<0.05),TC(β = 0.20,P <0.05),TG(β = 0.16,P <0.05),free CEL was associated with Hb A1c(β = 0.30,P <0.05),fasting blood glucose(β = 0.55,P <0.05),LDL-C(β = 0.54,P <0.05),TC(β = 0.86,P < 0.05),TG(β = 0.76,P <0.05),free MG-H1 was related to LDL-C(β = 0.21,P <0.05),TC(β = 0.35,P < 0.05),TG(β = 0.25,P <0.05),proteinbound CML was related to fasting blood glucose(β = 0.65,P <0.05),LDL-C(β =0.33,P <0.05),HDL-C(β =0.21,P <0.05),TC(β =0.51,P <0.05),and protein-bound CEL was associated with HDL-C(β =0.09,P <0.05).No correlation between total AGEs and glucose and lipid metabolism indices was observed(P> 0.05).Whereas,SAF was associated with Hb A1c(β = 0.47,P <0.05),LDL-C(β = 0.52,P <0.05),and TC(β =0.83,P <0.05).Conclusion: To some extent,AGE Reader has the ability to predict the changes of glucose and lipid metabolism.Due to its portability and simplicity,AGE Reader has the application value of large-scale epidemiologic study.ELISA was used to evaluate AGEs exposure only at high levels of exposure(> 5 AU).Free AGEs and bound AGEs have different effects on the prediction of glucose and lipid metabolism,which showed complementary effects. |
PDF Full Text Request