A Prospective Cohort Study On The Association Between Lipid And Type 2 Diabetes Incidence

Objective The purpose of this study is to find out the association between blood lipid indexes and the risk of type 2 diabetes（T2D）,to compare its ability to predict the risk of T2D,and to explore the main dyslipidemia metabolism mode and its critical value that affect the occurrence of T2D.Ultimately provide scientific basis for diabetes prevention and blood lipid managementMethods Based on Jinchang cohort,a total of 30546 subjects without T2D at baseline were included in this prospective cohort study.Cox proportional hazard regression model was used to calculate the intensity between different blood lipid,their combination patterns and the risk of T2D after adjusting for confounding factors.The dose-response relationship between lipids and incidence of T2D were analyzed by categorical analysis and restricted cubic spline（RCS）.ROC curve was plotted to evaluate the value of each blood lipid profile index in predicting T2D.Youden’s index was used to judge the optimal cut-offs for each lipid.Results 1.In the total population,the incidence of T2D in abnormal TC,TG,HDL-C and LDL-C were 4.96%,5.28%,5.36%and 4.32%,which were higher than that in normal groups（P<0.05）.2.Abnormal TG,TC and HDL-C in the general population were risk factors for the development of T2D,with the adjusted HRs of 1.618（95%CI:1.423～1.841）,1.161（95%CI:1.006～1.339）and 1.393（95%CI:1.229～1.578）,respectively.There were no significant association between LDL-C and T2D incidence.There were also similar findings in male individuals,while only TG has statistical effect among female population.3.When the concentration of lipids were divided into four levels according to the quartile,adjusted Cox proportional hazards model showed that the T2D incidence raised with the increase of TC and TG levels in the general population(P_trend<0.001),and reached the maximum in the fourth quartile（Q4）,with the HRs of 1.396（95%CI:1.180～1.651）and2.467（95%CI:2.003～3.038）.However,the T2D incidence was decreased with the increasing of HDL-C,and reached the minimum in Q4,with the HR of 0.818（95%CI:0.671～0.997）.RCS analysis showed that there were a positive dose-response relationship between TC and TG and the risk of T2D in the total population(P_overall<0.001).On the contrary,with the increasing of HDL-C concentration,the risk of T2D has a downward trend(P_overall=0.039).For LDL-C,a U-shaped curve was observed(P_overall<0.001).4.The risk of T2D increased with the increase of the number of abnormal lipid metabolism(P_trend<0.001).The effect was highest when there were 3 kinds of abnormal aggregation of lipid metabolism.Adjusted Cox regression model showed that the HRs of T2D incidence were 1.874（95%CI:1.557-2.256）,1.969（95%CI:1.562-2.483）and 1.413（95%CI:1.031-1.936）for total population,males and females,respectively.When high TG,high TC and low HDL-C coexist,the incidence risk of T2D was higher than that of other combined modes,with HRs of 4.169（95%CI:2.789-6.230）,4.125（95%CI:2.602～6.540）and3.332（95%CI:1.275～8.708）,respectively.5.With 1-unit increment in Non-HDL-C,remaining cholesterol,TG/HDL-C,TC/HDL-C and Ty G index,the HRs for T2D incidence were 1.242（95%CI:1.162～1.329）,1.583（95%CI:1.448～1.731）,1.061（95%CI:1.038～1.085）,1.182（95%CI:1.125～1.242）and2.047（95%CI:1.867～2.244）in general population,respectively.However,the association between LDL-C/HDL-C and the risk of T2D was only found in males.There was a positive dose-response relationship between the above six lipid composite indexes and the risk of T2D(P_overall<0.001).6.There was the maximum AUC for TG（0.646,95%CI:0.640～0.651）to predict the incidence of T2D among the four lipids（TC,TG,HDL-C,LDL-C）.The optimal cut-offs for TC,TG,HDL-C and LDL-C for evaluating T2D among the total population were5.09mmol/L,1.67mmol/L,1.40mmol/L and 3.08mmol/L,and the cut-offs were different between males and females.When the prediction model combining TG,TC and HDL-C had improved predictive value（AUC:0.651,95%CI:0.645～0.656）.7.ROC curve analyses showed that Ty G index strongly predicted T2D with AUC of0.718（95%CI:0.713～0.723）for the total population.The optimal cut-offs for Non-HDL-C,remaining cholesterol,TG/HDL-C,TC/HDL-C,LDL-C/HDL-C and Ty G index in predicting T2D were 3.66mmol/L,0.3mmol/L,1.14,3.45,2.46 and 8.82.Conclusions 1.Dyslipidemia significantly increase the risk of T2D at varying degree.Serum TG and HDL-C are the most important risk factors for T2D incident.2.The incidence of T2D is positively correlated with the aggregation of dyslipidemia.The combined effect of TG,TC and HDL-C has the greatest impact on the incidence of T2D,and the predictive effect is higher than any single blood lipid index3.Non-HDL-C,lipid ratios and Ty G index have a significant relationship with incident T2D.Ty G index is the most powerful marker for predicting the incidence of T2D.