Association Between Blood Glucose Level And Prognosis In Patients With Acute Heart Failure

BackgroundElevated levels of plasma glucose at hospital admission(acute hyperglycemia)common in patients with acute heart failure(AHF).It has been reported that acute hyperglycemia increase mortality in AHF patients,but previous studies have not been clear about the association between admission glucose(AG)and mortality in patients with AHF.Most of the previous studies have focused on patients with significant hyperglycemia or limited to small sample sizes.There is no large-scale prospective study on the association between AG levels and prognosis in patients with AHF.What's more,similar studies have not been reported on the long-term prognostic value of AG levels in China.Although many studies have shown that acute hyperglycemia can bring adverse prognosis.It suggests that we should strictly control the blood sugar in AHF patients.However,the goal of blood glucose control has not been defined,and the guidelines do not include acute hyperglycemia as a risk factor for risk stratification.It reflects that the association between acute hyperglycemia and mortality in AHF patients has not been properly established.Different AG levels and different grouping methods need to be further evaluated.In addition,diabetes mellitus(DM)is one of the important concomitant diseases in AHF patients.Studies have shown that DM is a risk factor for cardiovascular adverse outcome in AHF patients during hospitalization.It is not clear whether the effect of the similar AG level on the prognosis of DM and non-DM is different.Therefore,in an attempt to solve these problems,we conducted this study.Part 1:The impact of different AG levels on the prognosis of AHF patients.Part 2:The effect of DM history and AG level on the prognosis of AHF patients.Part 3:The effect of different grouping methods of AG levels on the prognosis of AHF patients.Part 1:The impact of different AG levels on the prognosis of AHF patientsObjectives To investigate the association between different AG levels and short-term and long-term mortality in DM or non-DM in AHF patients.Methods 512 consecutive patients hospitalized with AHF were included from June 2012 to December 2016,according to the AG level,were divided into the normal AG group(AG<7.53mmol/L)and the AG elevation Group(AG>7.53mmol/l),and the 2 groups were retrospectively studied.The median follow-up time was 20.2 months,the follow-up rate was 84%(430 cases).Results(?)In this study,the median blood glucose was 7.53 mmol/L,and the hyperglycemia patients was 49.2%.(?)The normal AG group patients was more common with NYHA-? class and has a higher serum sodium,while the AG elevation group was more commonly with NYHA-IV class and the history of AHF,a higher average age,faster heart rate and breathing.(?)AG level was significantly correlated with total cholesterol,low density lipoprotein cholesterol,mortality risk score(30-day,1-year).(?)Among the 8 concomitant diseases listed in this study,the proportion of hypertension was the highest(77.3%),and that of hyperglycemia group was more accompanied with DM and cerebral apoplexy.(?)The patients used diuretics,digitalis and vasoactive drugs was 80.7%,57.0%,55.5%respectively.Vasoactive drugs,insulin,ACEI/ARB,statins,oral antidiabetic drug are more commonly used in patients with acute hyperglycemia.(?)The average hospitalization of all patients was 11.39 days,the all-cause mortality of hospitalization was 16%,there was no difference between two groups(14.2%vs 17.9%,p = 0.280).The total mortality rate for all patients on the 30-day,1-year and 5-year was:21.5%,36.9%,49.2%respectively.The risk of death was higher in the group with elevated AG(30-day,1-year).There was no significant difference in 5-year all-cause mortality(P>0.05),and there was no difference in long-term prognosis between the two groups(P = 0.515).(?)The univariate regression analysis showed that the mortality risk score(30-day,1-year),age,NYHA grade,history of AHF,stroke,systolic blood pressure,blood sodium,C-reactive protein,albumin,were independent predictors of mortality in AHF patients,and multivariate regression analysis suggested that NYHA grade and the 30-day mortality risk score was a risk factor for all-causes mortality of AHF patients.Conclusions Nearly half of AHF patients have acute hyperglycemia in hospital.We found that there was no significant association between AG level and short-and long-term mortality in AHF patients.Part 2:The effect of diabetes history and AG level on the prognosisof AHF patientsObjectives To explore the effect of DM history,different AG levels and their interactions on the prognosis of AHF patients.Methods The AHF patients included in the first part(512 cases)were divided into DM group(180 cases)and non-DM group(332 cases)by whether there was a known history of DM.In each group,patients was divided into two groups according to the first part of the dichotomic method.The association between the different AG levels and the mortality on hospitalization,30-day,1-year,5-year were analyzed.In addiction,we compare difference of the DM history in COX model with different AG levels.The prognostic value of DM history,different AG levels and their interactions on the prognosis of AHF patients were compared.Results(?)The univariate analysis showed that there were no differences in hospitalization and the mortality on hospitalization,30-day,1-year,5-year among subgroups DM or non-DM and AG levels in DM group.(?)There was no significant difference in long-term cumulative mortality among the DM or non-DM,long-term cumulative mortality among groups with normal AG,and long-term cumulative mortality among groups with elevated AG(log-rank P>0.05).(?)The univariate regression analysis and multivariate regression analysis did not indicate that AG was a risk factor or predictor of all-cause mortality in AHF patients.Conclusions Regardless of the history of DM,there was no significant association between elevated AG levels and the mortality on hospitalization,30-day,1-year,5-year.AG is not a risk factor or predictor of all-cause mortality in AHF patients.Part 3:The effect of different grouping methods of AG levels on theprognosis of AHF patientsObjectives To compare the effect of different AG grouping methods on the prognosis of AHF patients.Methods The AHF patients included in the first part(512 cases)were divided into four groups by using the quartile method of AG level.To compare and analyze the association between the AG levels and the mortality on hospitalization,30-day,1-year,5-year under this grouping method.The differences in COX models with different AG levels were compared.Then the two groups used in the first part were compared directly with those in the quartile method.To observe the effect of different AG grouping methods on prognosis.Results(i)The univariate analysis showed:AHF patients were divided into four groups(F1?5.76mmol/l,F2:5.77-7.53mmol/l,F3:7.54-9.81mmol/l,F4>9.81mmol/l)using quartile method.Among the four groups,there were no differences in hospitalization and the mortality on hospitalization,30-day,1-year,5-year(P>0.05).There was no difference in the long-term prognosis between the four groups(Log-rank P = 0.884).(ii)Neither univariate regression nor multivariate regression analysis showed that AG was a risk factor or predictor of all-cause death in AHF patients.Conclusions(i)Using the quartile method of AG,there was still no significant association between AG and the mortality on hospitalization,30-day,1-year,5-year.AG was still not a risk factor or predictive factor for all-causes in AHF patients.(ii)The grouping method had no effect on the prognosis of AHF patients by AG.Final conclusions:Based on the above study,we found that:(?)The increase in AG was not related to the short-and long-term mortality in AHF patients.(?)The association between AG and the mortality on hospitalization,30-day,1-year,5-year was not significantly affected by different AG grouping methods and the history of DM.(?)AG was not a risk factor or predictive factor for all-causes mortality of AHF patients,was not significantly affected by different AG grouping methods and the history of DM.This conclusion remains to be confirmed by a larger prospective study.