Correlation Between The Variation Rate Of Triglyceride Glucose Index And Mortality In Critically Ⅲ Patients

Background and objective: Triglyceride-glucose index(TyG),namely ln [triglyceride(mg/dl)* plasma glucose(mg/dl)/ 2],is a new tool to indirectly reflect the levels of insulin resistance.Compared with the gold standard high insulin-normal glucose clamp experiment(HIEC),it has the advantages of cheap,easy to operate and easy to obtain repeatedly,which is not only proved to have a strong correlation with the measurement results of HIEC,but also found to reflect the condition and prognosis of chronic metabolic diseases to a certain extent.However,as a single-time index of responsetransient insulin resistance,it seems to have an innate defect in responding to changes in the patient’s condition.Severe patients are characterized by complex and variable conditions,high mortality rate and long hospital stay.At the same time,with the aging of the population,the structure of severe patients will be dominated by the elderly who are more likely to merge with chronic metabolic diseases.Therefore,this study innovatively proposed a new index triglyceride-glucose index variation rate(TyGVR),which was calculated by(last triglyceride glucose index-previous triglyceride glucose index)/ previous triglyceride glucose index.It mainly observes the changes of TyG,which not only makes full use of the advantages of TyG as cheap,easy operation,easy to obtain repeatedly,but also makes up for its dynamic defects to a certain extent.It can also bring help to the diagnosis and treatment of severe patients,and proves its value with a variety of statistical analysis.Methods: A total of 8,392 patients were included using the eligible clinical and general data in the fourth largest version of the Intensive Care Medical Information Market.First this study compared baseline data for all included patients and subsequently further grouped using propensity score matching.Survival analysis and restricted threeway regression analysis observed the correlation of TyG and TyGVR on hospital death and death within one year.The receiver operating characteristic curve(ROC)and AUC(Area under ROC)based on ROC curve were further used to analyze the value of TyG and TyGVR.Finally,we also conducted a further stratified analysis of multiple other potential factors of patients within each group through the forest map to observe the effect of these factors on the predictive value of TyG and TyG.Results: By comparing baseline data,patients with higher TyG and TyG VR were more likely to develop more severe metabolic abnormalities and related diseases,as well as to develop poor outcomes including prolonged hospitalization,prolonged hospital stay,invasive ventilation,and death.Survival analysis found that TyGVR before and after PSM was positively associated with mortality within one year.The mortality within one year was 30.86%,36.28%,37.82%,42.33%(before PSM),31.07%,35.13%,36.83% and 41.08%(after PSM),respectively.There was no statistical difference between TyG and patients within one year.The restricted cubic spline regression analysis found statistically significant differences in hospital death between TyGVR and patients within one year,and drew the corresponding relationship,while TyG failed to observe statistical differences in hospital death and within one year.ROC plots of six commonly used severe patient scores and ROC of hospital death and death within one year after combining TyG and TyGVR,comparing AUC found that all six severe scores had statistically significant AUC after combining TyG and TyGVR.The forest map comparison found that TyG predicted more hospital death in obesity and cerebrovascular disease,TyGVR predicted more hospital death for cerebrovascular disease and chronic kidney disease in patients,TyG predicted more death within one year in patients older than 65 years,non-diabetes,obesity and cerebrovascular disease,and TyGVR was more significant for death within one year in patients without chronic kidney disease.Conclusion: 1.TyGVR is an independent risk predictor of all-cause death in severe patients,and has predictive value for death in hospital or within one year.2 Both TyGVR and TyG can be used as complementary factors to currently commonly used severe patient scores,increasing their predictive value.3.TyGVR has a higher predictive value for mortality in severe patients without cerebrovascular disease and without chronic kidney disease.4.TyG has a parallel relationship with the current severity of severe patients.