Skip to main content
Download PDF

Prognostic significance of the stress hyperglycemia ratio and admission blood glucose in diabetic and nondiabetic patients with spontaneous intracerebral hemorrhage

Diabetology & Metabolic Syndrome volume 16, Article number: 58 (2024) Cite this article

Abstract

Background

The role of stress hyperglycemia ratio (SHR) on the prognosis of spontaneous intracerebral hemorrhage (ICH) in patients with different diabetic status has not been elucidated. This study aimed to evaluate the prognostic value of SHR and admission blood glucose (ABG) for the short- and long-term mortality in diabetic and nondiabetic populations with ICH.

Method

Participants with ICH were retrospectively retrieved from the Medical Information Mart for Intensive Care (MIMIC-IV). The primary outcome was all-cause 30-day and 1-year mortality. The association of SHR and ABG with the primary outcomes in diabetic and nondiabetic cohorts were assessed by Cox proportional hazard regression.

Results

Overall, 1029 patients with a median age of 71.09 (IQR: 60.05–81.97) were included. Among them, 548 (53%) individuals were male, and 95 (19%) as well as 323 (31%) ones experienced the 30-day and 1-year mortality, respectively. After adjusting for confounding variables, individuals in quintile 5 of SHR had significantly higher risk of the 30-day and 1-year mortality than those in quintile 1 in the whole cohort (30-day mortality: HR 3.33, 95%CI 2.01–5.51; 1-year mortality: HR 2.09, 95% CI 1.46-3.00) and in nondiabetic patients (30-day mortality: HR 4.55, 95%CI 2.33–8.88; 1-year mortality: HR 3.06, 95%CI 1.93–4.86), but no significant difference was observed in diabetic patients. Similar results were observed for ABG as a categorical variable. As continuous variable, SHR was independently correlated with the 30-day and 1-year mortality in both of the diabetic and nondiabetic cohorts (30-day mortality: HR 2.63, 95%CI 1.50–4.60. 1-year mortality: HR 2.12, 95%CI 1.33–3.39), but this correlation was only observed in nondiabetic cohort for ABG (HR 1.00, 95%CI 0.99–1.01 for both of the 30-day and 1-year mortality). Moreover, compared with ABG, SHR can better improve the C-statistics of the original models regarding the 30-day and 1-year outcomes, especially in patients with diabetes (p < 0.001 in all models).

Conclusion

SHR might be a more useful and reliable marker than ABG for prognostic prediction and risk stratification in critically ill patients with ICH, especially in those with diabetes.

Introduction

Spontaneous, nontraumatic, intracerebral hemorrhage (ICH) accounts for about 10–20% of all types of strokes which contributes the highest morbidity and mortality [1]. Despite much efforts made in the past decades, there are still few therapeutic options to improve prognosis after ICH [2]. Therefore, identifying modifiable factors that influence the outcome of ICH is important for that it may provide new targets for treatment.

Hyperglycemia at admission has been reported to be associated with poor outcomes in patients admitted with ICH [3,4,5]. The American Stroke Association guidelines also state that serum glucose should be monitored and hyperglycemia should be avoided in patients with ICH (Class I; Level of Evidence C) [6]. However, the admission blood glucose (ABG) level cannot entirely reflect the acute hyperglycemic state, which is also affected by the chronic glucose level. Therefore, to identify acute hyperglycemia more accurately, the stress hyperglycemia ratio (SHR) that combined ABG and glycosylated hemoglobin A1c (representing glycemic status of the past 2 ~ 3 months) was proposed. Since then, investigators have verified the predictive value of SHR on the prognosis of ACS and AMI [7,8,9,10]. More recently, the association between SHR and the higher risk of poor outcomes in patients with ICH was also reported [11, 12]. Nevertheless, as previous studies revealed that ABG and hyperglycemia was positively associated with short- and long-term poor outcomes in non-diabetic, but not diabetic individuals with ICH [13, 14], it is essential to clarify whether the impact of SHR on the prognosis of ICH is also dependent on different status of glucose metabolism.

In this study, we aimed to compare the predictive value of two measures of hyperglycemia (SHR and ABG) on the short- and long-term all-cause mortality in critically ill patients with ICH and in subgroups divided by diabetes mellitus, which may be useful to stratify and recognize individuals at high risk of all-cause mortality for timely healthcare management.

Method

Data sources

We conducted a retrospective study by retrieving data from the Medical Information Mart for Intensive Care (MIMIC-IV version 2.2) database [15]. MIMIC-IV is consisted of de-identified health-related data of more than 190,000 patients admitting to the intensive critical care units of the Beth Israel Deaconess Medical Center between 2008 and 2019. Since MIMIC-IV is de-identified to remove patients’ information and our study is centered around the analysis of this accessible database, which had already received the approval of an institutional review board (IRB), the requirement for patient informed consent and ethics approval is not indispensable.

Study population and data extraction

PostgresSQL (version 13.7.2) and Navicate Premium (version 16) was used to run structure query language (SQL) and extract data from MIMIC-IV. One author (Yang Yang) received proper authorization to access the database (Record ID: 48,776,647), and was responsible for data extraction. We included adult patients (> 18 years of age) with a diagnosis of ICH according to the International Classification of Diseases, 9th and 10th Revision (ICD9 and ICD10). The exclusion criteria were as follows: patients with ICH (1) dying or leaving within 24 h since intensive care unit (ICU) admission; (2) with multiple ICU admissions, for whom only the data of the first admission due to ICH were extracted; (3) with missing ABG and HbA1c data on the first day of admission. Finally, a total of 1028 patients were extracted and enrolled in this study.

For the final cohort, we collected the following data: (1) age and gender; (2) comorbidity as coded and defined by the ICD-9 and ICD-10 (Additional file 1: Table S1); (3) severity of illness scores on the first day of ICU admission, which include acute physiology score III (APSIII), simplified acute physiological score II (SAPSII), sequential organ failure assessment (SOFA), and oxford acute severity of illness score (OASIS); (4) the first result of glucose and HbA1c since ICU admission; (5) the average laboratory results on the first day of ICU admission; (6) the mean values of vital signs during the first 24 h of ICU stay; (7) medication data during ICU stay including usage of vasoactive drugs, renal replacement treatment (RRT) and invasive ventilation.

Diabetes was diagnosed according to the definition of the American Diabetes Association as (1) a definite history of diabetes, or (2) HbA1c value ≥ 6.5% [16]. The value of SHR was calculated as follows: SHR = ABG(mg/dL)/(28.7 × HbA1c(%) − 46.7) [17]. Then, the final cohort was divided into five groups according to the quintile of the SHR. (Fig. 1).

Fig. 1
figure 1

Flowchart illustrating the screening process of critically ill patients with ICH from the MIMIC-IV database

Full size image

Clinical outcomes

The primary outcomes were all-cause 30-day and 1-year mortality. It is worth noting that the MIMIC-IV database restricts access to follow-up dates beyond one year from the last hospital discharge. Therefore, no patient was lost to follow-up within 1 year, but the database does not facilitate insights into patient mortality beyond the 1-year timeframe.

Statistical analysis

The Shapiro-Wilk test was used to assess the distribution of continuous variables. Data were expressed as mean ± standard deviation (SD) for normally distributed variables, median (interquartile ranges) for non-normal data, and number (percentages) for categorical data. Normally distributed variables were compared by unpaired Student’ t test or ANOVA, and variables with a skewed distribution were compared by Mann-Whitney U-test or Kruskal-Wallis test. The chi-squared test or Fisher’s exact test were performed to assess the differences in categorial variables between groups.

Kaplan-Meier’s survival analysis combined with Log-rank test was then conducted to visualize and compare the probability of the 30-day and 1-year death among groups divided by quintile of the SHR and ABG. Before Cox regression analysis was conducted, the association of SHR and ABG with other continuous variables were assessed by Pearson’s correlation test or Spearman’s rank correlation test (Additional file 2: Figure S1). Univariable and multivariable Cox proportional hazard regression was then performed to evaluate the impact of SHR and ABG on the 30-day and 1-year outcomes. In multivariable analysis, we adjusted the variables related to the 30-day or 1-year mortality in univariate Cox regression with a p value < 0.01. Further, we used a restricted cubic spline regression model with four knots to evaluate the nonlinear association of SHR and ABG with the primary outcomes. Then, Harrell’s C-statistic was calculated to assess whether adding SHR or ABG into a model of the severity of illness scores (APSIII, SAPSII, SOFA, and OASIS) could improve the discrimination of the model for the 30-day and 1-year mortality. Finally, subgroup analyses were conducted by stratifying patients according to their diabetic status.

We use R software (version 4.3.1) for data analysis. “VIM” package of R was used to visualize the distribution of missing values. All of the collected variables had missing ratio less than 20% (Additional file 3: Figure S2). The “mice” package of R was adopted to address missing values by multiple imputation to obtain 5 imputation datasets in the process of Cox regression. Besides, the “corrplot” package was used to visualize the associations of SHR, ABG and HbA1c with continuous variables. “survminer” packages was used to conduct Kaplan-Meier’s survival analysis. “survivalROC” packages was used to estimate the cut-off values. A two-tailed p value of < 0.05 was considered statistically significant.

Results

Baseline characteristics

The median age of the whole cohort was 71.09 (IQR: 60.05–81.97) years and among them, 548 (53%) patients were male and 304 (30%) ones had diabetes. The 30-day and 1-year mortality of the study cohort were 19% and 31%, respectively. Besides, an ascending gradient with respect to the severity of illness scores, prevalence of diabetes, values of mean heart rate and respiratory rate, number of lung infection, ICU and hospital stay time, as well as the 30-day and 1-year mortality, whereas a descending gradient regarding value of mean arterial pressure was found from quintile 1 to quintile 5 of SHR (Table 1). Furthermore, patients with diabetes had higher score of APSIII, higher prevalence of congestive heart failure, renal diseases and hypertension, lower value of white blood cell count, and higher levels of ABG and HbA1c. While no difference was observed for the other variables between the two groups (Additional file 4: Table S2).

Table 1 Baseline characteristics of participants grouped by quintile of SHR a,b
Full size table

Association between SHR, ABG and the 30-day mortality

The levels of SHR and ABG in the 30-day non-survivors was significantly higher than that in the 30-day survivors (1.18 (IQR:1.00-1.47) vs. 1.01 (IQR:0.88–1.19), p < 0.001 for SHR; 139 (IQR:117-178.5) vs.121 (IQR:102–150), p < 0.001 for ABG) (Fig. 2). The Kaplan-Meier’s survival analysis combined with log-rank test exhibited significant difference in the 30-day mortality among patients divided by the quintile of SHR and ABG (p < 0.001 for both). The comparison between groups showed that the 30-day mortality was significantly higher in patients with an SHR in the fifth interval (Q5) than those in the other four intervals (Q1 ~ Q4 vs. Q5: p < 0.001), but no difference was observed between patients in Q5 and those in Q3 and Q4 of ABG (Q1 ~ Q2 vs. Q5: p < 0.001; Q3 vs. Q5: p = 0.07; Q4 vs. Q5: p = 0.22) (Fig. 3).

Fig. 2
figure 2

The boxplot of SHR and ABG stratified by primary outcomes. (a)~(c) The level of SHR and ABG in the 30-day survivors and non-survivors. (d)~(f) The level of SHR and ABG in the 1-year survivors and non-survivors

Full size image
Fig. 3
figure 3

The Kaplan-Meier’s survival curve estimated of critically ill patients with ICH grouped by the quintile of SHR and ABG. (a) the 30-day survival analysis of patients divided by the quartiles of SHR (Log-rank p < 0.0001). (b) the 30-day survival analysis of patients divided by the quartiles of ABG (Log-rank p < 0.0001). (c) the 1-year survival analysis of patients divided by the quartiles of SHR (Log-rank p < 0.0001). (d) the 1-year survival analysis of patients divided by the quartiles of ABG (Log-rank p < 0.0001)

Full size image

In the adjusted model of Cox regression analysis, patients in Q4 and Q5 of SHR had significantly higher risk of the 30-day death compared with those in Q1 (Q4 vs. Q1: adjusted HR 1.88, 95%CI 1.08–3.27; Q5 vs. Q1: adjusted HR 3.33, 95%CI 2.01–5.51), but no significant difference was observed between patients in Q4 and Q1 of ABG (Q4 vs. Q1: adjusted HR 1.58, 95% CI 0.96–2.59; Q5 vs. Q1: adjusted HR 2.21, 95%CI 1.37–3.57) (Table 2). Furthermore, restricted cubic spline analysis after fully adjusted demonstrated a nearly linear association of SHR and ABG with the risk of 30-day mortality (p for non-linear = 0.054, p for overall < 0.001 for SHR; p for non-linear = 0.511, p for overall < 0.001 for ABG) (Fig. 4).

Table 2 Hazard ratio (HR) for all-cause 30-day or 1-year mortality of SHR, ABG and HbA1C in the whole cohort
Full size table
Fig. 4
figure 4

Restricted cubic spline curve for the Hazard ratio of SHR and ABG. SHR 1.04 and ABG 124, represented by the vertical dotted lines, was an estimated hazard ratio of 1.0. The horizontal dotted line represents the hazard ratio of 1.0. (a) Restricted cubic spline of SHR for the 30-day mortality. (b) Restricted cubic spline of ABG for the 30-day mortality. (c) Restricted cubic spline of SHR for the 1-year mortality. (d) Restricted cubic spline of ABG for the 1-year mortality

Full size image

Association between SHR, ABG and the 1-year mortality

The levels of SHR and ABG in the 1-year non-survivors was significantly higher than that in the 1-year survivors (1.11 (IQR:0.94–1.38) vs. 1.01 (IQR:0.88–1.19), p < 0.001 for SHR; 135 (IQR:110-169.5) vs. 120 (IQR:102–149), p < 0.001 for ABG) (Fig. 2). The Kaplan-Meier’s survival analysis combined with log-rank test showed significant difference in the 1-year survival among patients divided by the quintile of SHR and ABG (p < 0.001 for both). The comparison between groups showed that patients’ 1-year mortality was significantly higher in Q5 of SHR than that in the other four intervals (Q1 ~ Q4 vs. Q5: p < 0.001), but no significant difference was observed between patients in Q5 and Q4 of ABG (Q1 ~ Q2 vs. Q5: p < 0.001, Q3 vs. Q5: p = 0.003, Q4 vs. Q5: p = 0.22 for ABG) (Fig. 3).

In the adjusted model of Cox regression analysis, patients in Q5 of SHR and ABG had significantly higher risk of the 1-year mortality compared with those in Q1 (adjusted HR 2.09, 95% CI 1.46-3.00 for SHR; adjusted HR 1.79, 95% CI 1.26–2.54 for ABG) (Table 2). Moreover, restricted cubic spline analysis after fully adjusted demonstrated a nearly linear association of SHR and ABG with the risk of 1-year mortality (p for non-linear = 0.287, p for overall < 0.001 for SHR; p for non-linear = 0.351, p for overall = 0.001 for ABG) (Fig. 4).

Association of SHR and ABG with the primary outcomes in the diabetic and non-diabetic subgroups

To further assess the impact of glucose metabolism status on the risk stratification value of SHR and ABG for the 30-day and 1-year outcomes of ICH, Cox regression analysis was conducted in the diabetic and non-diabetic subgroups.

In the non-diabetic subgroup, patients in Q5 of SHR and ABG had significantly higher risk of 30-day and 1-year mortality compared with those in Q1 (30-day mortality: adjusted HR 4.55, 95%CI 2.33–8.88 for SHR; adjusted HR 2.71, 95%CI 1.55–4.73 for ABG. 1-year mortality: adjusted HR 3.06, 95%CI 1.93–4.86 for SHR; adjusted HR 2.38, 95%CI 1.56–3.63 for ABG). Moreover, significantly higher risk of 30-day and 1-year mortality was also existed in Q4 of SHR compared with Q1 (30-day mortality: adjusted HR 2.92, 95%CI 1.46–5.84. 1-year mortality: adjusted HR 2.02, 95%CI 1.24–3.28). (Additional file 5: Table S3).

In the diabetic subgroup, the risk of 30-day and 1-year mortality had no significant difference between patients in Q2 ~ Q5 of SHR and ABG and those in Q1. However, as continuous variable, SHR was significantly associated with the 30-day and 1-year mortality in the adjusted model, but that was not the case for ABG (30-day mortality: adjusted HR 2.63, 95%CI 1.50–4.60 for SHR; adjusted HR 1.00, 95%CI 0.99–1.01 for ABG. 1-year mortality: adjusted HR 2.12, 95%CI 1.33–3.39 for SHR; adjusted HR 1.00, 95%CI 0.99–1.01 for ABG) (Table 3).

Table 3 Hazard ratio (HR) for all-cause 30-day or 1-year mortality of SHR and ABG in the patients with diabetes
Full size table

Predictive value of SHR and ABG for the primary outcomes

Compared with the original Cox regression models of the severity of illness scores (APSIII, SAPSII, SOFA, or OASIS), adding SHR or ABG into the original models can significantly increase the C-statistic for predicting both of the 30-day and 1-year outcomes in the whole cohort and in patients without diabetes (Table 4; Additional file 6: Table S4). In patients with diabetes, adding SHR into the original models can significantly increase the C-statistic for predicting the 30-day outcome, while in terms of the 1-year prognosis, SHR can improve the predictive efficiency of the original model constructed by SOFA or OASIS. Notably, no significant improvement of predictive efficiency for both of the 30-day and 1-year outcomes was observed when adding ABG into anyone of the original models (Additional file 7: Table S5). More importantly, compared with ABG, SHR exhibited higher predictive efficiency and can better improve the C-statistics of the original models regarding 30-day and 1-year outcomes, especially in patients with diabetes (Additional file 8: Table S6; Additional file 9: Table S7).

Table 4 Discrimination ability of SHR and ABG for all cause 30-day and 1-year mortality in the whole cohort
Full size table

Discussion

Main findings

This study, for the first time, compared the prognostic effect and predictive value of ABG and SHR for the short- and long-term outcomes in critically ill patients with ICH, and two main findings have been obtained. First, SHR, instead of ABG, was an independent risk factor of the 30-day and 1-year mortality of ICH in the diabetic population. Second, SHR might exhibited higher predictive efficiency for the 30-day and 1-year mortality of ICH than ABG, especially in the diabetic cohort.

Findings of previous studies and contribution of this study

Since an appropriate predictor should be effectively intervened through existing management to improve the outcome, it is important to determine whether elevated ABG after ICH is associated with increased poor outcomes. Evidence from two large-scale, multi-center studies supports ABG as an independent predictor of poor outcome in patients with ICH [5, 18], but the prognostic effects of ABG based on patients’ diabetic status is not elucidated in these studies due to the lack of direct comparison between diabetics and non-diabetics. Based on this deficiency, Sun et al. conducted a multi-center, prospective cohort study which indicated that elevated ABG is an independent predictor of the 3-month poor outcome in ICH patients and its prognostic value is greater in non-diabetics than diabetics [14]. Besides, an earlier study also found that high ABG in nondiabetics was a significant predictor of death during the first 28 days of the onset of ICH [19]. Consistent with these findings, our study proved that ABG was positively associated with higher 30-day and 1-year mortality in the whole study cohort with ICH and in the non-diabetic subgroup.

A predominant defect that may restrict the clinical practice of ABG is that the level of ABG may be influenced by a number of factors, such as diabetic status, recent glycemia control status, etc. Therefore, the rise of blood glucose due to ICH compared with the background status may be a better predictor of the outcomes of ICH. However, the level of blood glucose prior to ICH is nearly impossible to obtain, especially in non-diabetic patients. Fortunately, this thorny problem is expected to be overcome by using SHR, an index that more accurately reflects the extent of stress-induced hyperglycemia by correction for chronic glycemic status [20]. A pile of studies had demonstrated that SHR is a useful indicator for predicting the outcomes of acute myocardial infarction [7, 10, 21,22,23]. Moreover, in a landmark study on the association between SHR and the risk of adverse outcomes in acute ischemic stroke (AIS), Huang et al. incorporated a total of thirteen studies encompassing 184,179 individuals and found that higher SHR was significantly associated with increased risk of various adverse outcomes, including a 2.64-fold increased risk of 3-month poor functional outcomes, a 3.11-fold increased risk of 3 month mortality, and a 2.80-fold increased risk of 1-year mortality [24]. However, evidence about the role of SHR in the outcomes of ICH is still limited. Chu and his colleagues seemed to be the first to investigate the effect of SHR on ICH. They revealed that SHR is independently associated with hematoma expansion and 3-month mortality in ICH patients [11]. Consistent with their findings, our study proved that SHR was positively associated with the risk of short- and long-term mortality in the critically ill patients with ICH. More importantly, we proved that SHR, as a continuous variable, was independently related to the 30-day and 1-year mortality in ICH patients with diabetes, but that was not the case for ABG. This result was similar to a recent study derived from the Chinese Cerebral Hemorrhage: Mechanisms and Intervention Study (CHEERY) [25]. In this study, Chen et al. divided ICH patients into nondiabetic normoglycemia (NDN), diabetic normoglycemia (DN), diabetic hyperglycemia (DH), and stress-induced hyperglycemia (SIH) groups, and found that compared with patients with NDN, DH did not increase the risk of poor outcome and mortality, whereas SIH was an independent risk factor for pulmonary infection and 30- and 90-day death after ICH [25]. Although Chen et al. defined stress hyperglycemia by other method (having a diabetes history or HbA1c ≥ 6.5%, and admission blood glucose ≥ 7.8 mmol/L) [26] instead of SHR, the results from their study and ours indicated that stress hyperglycemia may be more closely associated with adverse prognosis in diabetic patients with ICH than admission hyperglycemia. In addition, we revealed that SHR had higher predictive value than ABG respect to the 30-day and 1-year outcomes, especially in patients with diabetes, further supporting SHR as a more reliable and useful indicator than ABG in the risk stratification and prognosis prediction of diabetic patients with ICH.

Detrimental effects of stress hyperglycemia and its mechanisms

The effects of stress hyperglycemia on ICH are complex and not fully understood. Stress hyperglycemia has been shown to promote the generation of oxygen-free radical, which can then induce toxic effect and lead to neuronal apoptosis in experimental ICH modle [27, 28]. It is reported that hyperglycemia after ICH is associated with neutrophil and lymphocyte ratio, indicating that stress hyperglycemia may induce inflammatory responses and then promote secondary brain injury. Moreover, studies have also shown that the superoxide production induced by hyperglycemic can disrupt the blood–brain barrier (BBB) and aggravate cerebral edema in rat models [29]. Hematoma expansion has been considered as an important contributor to the worsening prognosis of ICH, studies has revealed that stress hyperglycemia leads to hematoma expansion possibly by impairing vessels integrity nearby the site of initial bleeding and increasing the expression of nuclear factor kappa B and matrix metalloproteinase-9 [30]. Meanwhile, Aquaporin-4 (AQP4) is the most abundant aquaporin in the brain that protects neuron from apoptosis by alleviating BBB disruption and brain edema [31, 32]. Studies had demonstrated that hyperglycemia may exacerbate outcome of ICH through downregulating expression of AQP4 [33]. In addition, by using magnetic resonance imaging with spectroscopy, scholars had found an association of hyperglycemia and brain lactate with penumbral damage in patients with ischemic stroke, therefore it is reasonable to infer that there may be a similar injury mechanism in ICH [34].

Mechanisms of the different prognostic values of ABG

Our data exhibited different prognostic values of elevated ABG in ICH patients with and without diabetes, which was consistent with previous findings on ACS or AMI [8, 35]. It is difficult to clearly explain this between-group difference in the effect of ABG, but several possible mechanisms may contribute to it. First, undiagnosed diabetes or pre-diabetes may be existed in some non-diabetic patients with elevated ABG. This part of patients may represent a higher-risk cohort with poor outcomes, especially in those with glucose ≥ 11.1 mmol/L. Second, as studies had shown that insulin resistance is a marker of increased risk of incident ischemic stroke in the nondiabetic individuals [36, 37], it is reasonable to infer that insulin resistance may exist in non-diabetic patients with ICH, particular in those with elevated ABG. Compared with the diabetic patients with hyperglycemia during ICH, the non-diabetic ones may expose to a greater degree of fluctuation in insulin resistance in order to reach the same level of hyperglycemia, which may result in higher risk of adverse outcome. Third, the non-diabetics may suffer from greater degree of stress reactions to reach the same hyperglycemic state as their diabetic counterparts, which may lead to higher mortality in ICH patients without diabetes [38]. It is worth noting that we focused on critically ill patients in this study. In a state of severe illness, patients’ body can enter a maladaptive state of allostatic overload, which may amplify the adverse effects of hyperglycemia mentioned above [39]. Forth, elevated ABG in the non-diabetic patients was always overlooked during hospitalization. A better management of elevated ABG in the diabetic patients compared with their non-diabetic counterparts may result in different outcomes.

Implications for glucose management in ICH

Glucose monitoring and management is often considered as a part of the general care of all patients, including those with ICH. In the NICE-SUGAR trial, a blood glucose target of < 180 mg/dL was associated with lower mortality than a target of 81 to 108 mg/dL, suggesting that targets for treating hyperglycemia should be less intensive in critically ill patients [40]. Mauro and colleagues reported that tight systemic glycemic control (80–110 mg/dL) may impair cerebral glucose metabolism after severe brain injury, which in turn correlates with increased mortality [41]. These evidences support a viewpoint that intensive insulin therapy, which has been proved to be beneficial in patients with ischemic myocardium, may increase the risk of hypoglycemic events and worsen outcomes in ICH patients. However, the optimal glucose level at which treatment should be initiated and the target range of treatment are still unclear. One reason for this confusion, which has been stated by the American Stroke Association, is that the relationship among serum glucose, the timing of that measurement, and the presence/absence of comorbid diabetes remains unclear [42,43,44,45]. Most of the previous studies take advantage of ABG or systemic glucose as the target of their intervention, but as the level of ABG or systemic glucose can be influenced by a number of factors, such as different diabetic status or insulin resistance levels, these two indicators may not be the good choice to reflect the acute glycemic status. In this case, SHR may be better to identify the true stress hyperglycemia by correction for chronic glycemic status. Evidence from this study also supports SHR as a better prognostic indicator for patients with ICH, especially for those with diabetes. Therefore, a more personalized strategy that based on patients’ diabetic status and combined SHR and ABG as the target may be a new choice for glucose management in patients with ICH. But this idea needs to be further confirmed in the future by high-quality randomized controlled studies.

Limitations of the study

Several limitations of this study should be concerned. First, the retrospective nature of the observational study determined that although multivariable-adjusted analysis was conducted, unmeasured variables could impact outcomes. Second, we excluded ICH patients with missing data of ABG and HbA1c, but if the pattern of missing is not completely random, selection bias may compromise the reliability of the conclusions. Third, as the value of admission glucose is also subjected to meal timing, fasting plasma glucose may be superior to reflect the true state of blood glucose, but this indicator cannot be collected from the MIMIC-IV database. Fourth, both SHR and ABG are derived from the first result of blood glucose test since ICU admission, which cannot fully reflect the profile of glucose fluctuation in the acute phase of ICH. Therefore, the association between glucose or SHR variation identified by continuous glucose monitoring and prognosis of ICH need to be evaluated in the future. Last but not least, as a single-center study, the conclusions of this study cannot be directly extrapolated to general population without validating in more centers.

Conclusion

This study demonstrated that SHR might be a more useful and reliable marker than ABG for prognostic prediction and risk stratification in critically ill patients with ICH, especially in those without diabetes. Further prospective studies with larger population are needed to confirm the findings and determine whether glucose management targeting SHR will improve clinical outcomes of critically ill patients with ICH.

Data availability

The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.

Abbreviations

ABG:

Admission blood glucose

ACS:

Acute coronary syndrome

ALT:

Alanine aminotransferase

AMI:

Acute myocardial infarction

APSIII:

Acute physiology score III

APTT:

Activated partial thromboplastin time

BUN:

Blood urea nitrogen

GCS:

Glasgow coma scale

HbA1c:

Hemoglobin A1c

ICD:

International Classification of Diseases

ICH:

Intracerebral hemorrhage

MI:

Myocardial infarct

MIMIC-IV:

The Medical Information Mart for Intensive Care

OASIS:

Oxford acute severity of illness score

PT:

Prothrombin time

RDW:

Red cell distribution width

RRT:

Renal replacement treatment

SAPSII:

Simplified acute physiological score II

SHR:

Stress hyperglycemia ratio

SOFA:

Sequential organ failure assessment

VIF:

Variance inflation factor

WBC:

White blood cell count

References

  1. Cordonnier C, Demchuk A, Ziai W, Anderson C. Intracerebral haemorrhage: current approaches to acute management. Lancet (London England). 2018;392(10154):1257–68.

    Article  PubMed  Google Scholar 

  2. Keep R, Hua Y, Xi G. Intracerebral haemorrhage: mechanisms of injury and therapeutic targets. Lancet Neurol. 2012;11(8):720–31.

    Article  PubMed  CAS  Google Scholar 

  3. Guo X, Li H, Zhang Z, et al. Hyperglycemia and mortality risk in patients with primary intracerebral hemorrhage: a Meta-analysis. Mol Neurobiol. 2016;53(4):2269–75.

    Article  PubMed  CAS  Google Scholar 

  4. Tan X, He J, Li L, et al. Early hyperglycaemia and the early-term death in patients with spontaneous intracerebral haemorrhage: a meta-analysis. Intern Med J. 2014;44(3):254–60.

    Article  PubMed  CAS  Google Scholar 

  5. Saxena A, Anderson C, Wang X, et al. Prognostic significance of hyperglycemia in Acute Intracerebral Hemorrhage: the INTERACT2 study. Stroke. 2016;47(3):682–8.

    Article  PubMed  CAS  Google Scholar 

  6. Greenberg S, Ziai W, Cordonnier C, et al. 2022 Guideline for the management of patients with spontaneous intracerebral hemorrhage: a Guideline from the American Heart Association/American Stroke Association. Stroke. 2022;53(7):e282–e361.

    Article  PubMed  CAS  Google Scholar 

  7. Marenzi G, Cosentino N, Milazzo V, et al. Prognostic value of the Acute-to-chronic glycemic ratio at Admission in Acute myocardial infarction: a prospective study. Diabetes Care. 2018;41(4):847–53.

    Article  PubMed  CAS  Google Scholar 

  8. Yang Y, Kim T, Yoon K, et al. The stress hyperglycemia ratio, an index of relative hyperglycemia, as a predictor of clinical outcomes after percutaneous coronary intervention. Int J Cardiol. 2017;241:57–63.

    Article  PubMed  Google Scholar 

  9. Şimşek B, Çınar T, Tanık V, et al. The association of acute–to–chronic glycemic ratio with no-reflow in patients with ST–segment elevation myocardial infarction undergoing primary percutaneous coronary intervention. Kardiologia Polska. 2021;79(2):170–8.

    Article  PubMed  Google Scholar 

  10. Yang J, Zheng Y, Li C, et al. The impact of the stress hyperglycemia ratio on short-term and long-term poor prognosis in patients with Acute Coronary Syndrome: insight from a large cohort study in Asia. Diabetes Care. 2022;45(4):947–56.

    Article  MathSciNet  PubMed  CAS  Google Scholar 

  11. Chu H, Huang C, Tang Y, Dong Q, Guo Q. The stress hyperglycemia ratio predicts early hematoma expansion and poor outcomes in patients with spontaneous intracerebral hemorrhage. Ther Adv Neurol Disord. 2022;15:17562864211070681.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  12. Li S, Wang Y, Wang W, Zhang Q, Wang A, Zhao X. Stress hyperglycemia is predictive of clinical outcomes in patients with spontaneous intracerebral hemorrhage. BMC Neurol. 2022;22(1):236.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  13. Qureshi A, Huang W, Lobanova I, et al. Effect of moderate and severe persistent hyperglycemia on outcomes in patients with Intracerebral Hemorrhage. Stroke. 2022;53(4):1226–34.

    Article  PubMed  CAS  Google Scholar 

  14. Sun S, Pan Y, Zhao X, et al. Prognostic value of admission blood glucose in Diabetic and non-diabetic patients with Intracerebral Hemorrhage. Sci Rep. 2016;6:32342.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  15. Johnson A, Bulgarelli L, Shen L, et al. MIMIC-IV, a freely accessible electronic health record dataset. Sci data. 2023;10(1):1.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  16. ElSayed N, Aleppo G, Aroda V et al. Addendum. 2. Classification and Diagnosis of Diabetes: Standards of Care in Diabetes-2023. Diabetes Care. 2023;46(Suppl. 1):S19-S40. Diabetes care 2023; 46(9): 1715.

  17. Roberts G, Quinn S, Valentine N, et al. Relative hyperglycemia, a marker of critical illness: introducing the stress hyperglycemia ratio. J Clin Endocrinol Metab. 2015;100(12):4490–7.

    Article  PubMed  CAS  Google Scholar 

  18. Passero S, Ciacci G, Ulivelli M. The influence of diabetes and hyperglycemia on clinical course after intracerebral hemorrhage. Neurology. 2003;61(10):1351–6.

    Article  PubMed  Google Scholar 

  19. Fogelholm R, Murros K, Rissanen A, Avikainen S. Admission blood glucose and short term survival in primary intracerebral haemorrhage: a population based study. J Neurol Neurosurg Psychiatry. 2005;76(3):349–53.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  20. Nathan D, Kuenen J, Borg R, Zheng H, Schoenfeld D, Heine R. Translating the A1C assay into estimated average glucose values. Diabetes Care. 2008;31(8):1473–8.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  21. Xu W, Yang Y, Zhu J, et al. Predictive value of the stress hyperglycemia ratio in patients with acute ST-segment elevation myocardial infarction: insights from a multi-center observational study. Cardiovasc Diabetol. 2022;21(1):48.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  22. Sia C, Chan M, Zheng H, et al. Optimal glucose, HbA1c, glucose-HbA1c ratio and stress-hyperglycaemia ratio cut-off values for predicting 1-year mortality in diabetic and non-diabetic acute myocardial infarction patients. Cardiovasc Diabetol. 2021;20(1):211.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  23. Cui K, Fu R, Yang J, et al. The impact of fasting stress hyperglycemia ratio, fasting plasma glucose and hemoglobin A1c on in-hospital mortality in patients with and without diabetes: findings from the China acute myocardial infarction registry. Cardiovasc Diabetol. 2023;22(1):165.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  24. Huang Y, Li Z, Yin X. Stress hyperglycemia and risk of adverse outcomes in patients with acute ischemic stroke: a systematic review and dose-response meta-analysis of cohort studies. Front Neurol. 2023;14:1219863.

    Article  PubMed  PubMed Central  Google Scholar 

  25. Chen S, Wan Y, Guo H, et al. Diabetic and stress-induced hyperglycemia in spontaneous intracerebral hemorrhage: a multicenter prospective cohort (CHEERY) study. CNS Neurosci Ther. 2023;29(4):979–87.

    Article  PubMed  Google Scholar 

  26. Moghissi E, Korytkowski M, DiNardo M, et al. American Association of Clinical Endocrinologists and American Diabetes Association consensus statement on inpatient glycemic control. Diabetes Care. 2009;32(6):1119–31.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Won S, Tang X, Suh S, Yenari M, Swanson R. Hyperglycemia promotes tissue plasminogen activator-induced hemorrhage by increasing superoxide production. Ann Neurol. 2011;70(4):583–90.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  28. Chiu C, Chen T, Chin L, et al. Investigation of the effect of hyperglycemia on intracerebral hemorrhage by proteomic approaches. Proteomics. 2012;12(1):113–23.

    Article  PubMed  CAS  Google Scholar 

  29. Song E, Chu K, Jeong S, et al. Hyperglycemia exacerbates brain edema and perihematomal cell death after intracerebral hemorrhage. Stroke. 2003;34(9):2215–20.

    Article  PubMed  CAS  Google Scholar 

  30. Zheng Y, Hu Q, Manaenko A, et al. 17β-Estradiol attenuates hematoma expansion through estrogen receptor α/silent information regulator 1/nuclear factor-kappa b pathway in hyperglycemic intracerebral hemorrhage mice. Stroke. 2015;46(2):485–91.

    Article  PubMed  CAS  Google Scholar 

  31. Chu H, Tang Y, Dong Q. Protection of Vascular endothelial growth factor to Brain Edema following intracerebral hemorrhage and its involved mechanisms: effect of Aquaporin-4. PLoS ONE. 2013;8(6):e66051.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  32. Chu H, Ding H, Tang Y, Dong Q. Erythropoietin protects against hemorrhagic blood-brain barrier disruption through the effects of aquaporin-4. Laboratory investigation. J Tech Methods Pathol. 2014;94(9):1042–53.

    CAS  Google Scholar 

  33. Chiu C, Chen C, Shen C, et al. Hyperglycemia exacerbates intracerebral hemorrhage via the downregulation of aquaporin-4: temporal assessment with magnetic resonance imaging. Stroke. 2013;44(6):1682–9.

    Article  PubMed  CAS  Google Scholar 

  34. Parsons M, Barber P, Desmond P, et al. Acute hyperglycemia adversely affects stroke outcome: a magnetic resonance imaging and spectroscopy study. Ann Neurol. 2002;52(1):20–8.

    Article  PubMed  Google Scholar 

  35. Xu W, Song Q, Wang X, et al. Association of stress hyperglycemia ratio and in-hospital mortality in patients with coronary artery disease: insights from a large cohort study. Cardiovasc Diabetol. 2022;21(1):217.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  36. Rundek T, Gardener H, Xu Q, et al. Insulin resistance and risk of ischemic stroke among nondiabetic individuals from the northern Manhattan study. Arch Neurol. 2010;67(10):1195–200.

    Article  PubMed  PubMed Central  Google Scholar 

  37. Kernan W, Inzucchi S, Viscoli C, Brass L, Bravata D, Horwitz R. Insulin resistance and risk for stroke. Neurology. 2002;59(6):809–15.

    Article  PubMed  CAS  Google Scholar 

  38. Jotic A, Milicic T, Covickovic Sternic N et al. Decreased Insulin Sensitivity and Impaired Fibrinolytic Activity in Type 2 Diabetes Patients and Nondiabetics with Ischemic Stroke. International journal of endocrinology. 2015; 2015: 934791.

  39. Brame A, Singer M. Stressing the obvious? An allostatic look at critical illness. Crit Care Med. 2010;38:600–7.

    Article  Google Scholar 

  40. Finfer S, Chittock D, Su S, et al. Intensive versus conventional glucose control in critically ill patients. N Engl J Med. 2009;360(13):1283–97.

    Article  PubMed  Google Scholar 

  41. Oddo M, Schmidt J, Carrera E, et al. Impact of tight glycemic control on cerebral glucose metabolism after severe brain injury: a microdialysis study. Crit Care Med. 2008;36(12):3233–8.

    Article  PubMed  CAS  Google Scholar 

  42. Béjot Y, Aboa-Eboulé C, Hervieu M, et al. The deleterious effect of admission hyperglycemia on survival and functional outcome in patients with intracerebral hemorrhage. Stroke. 2012;43(1):243–5.

    Article  PubMed  Google Scholar 

  43. Lee S, Kim B, Bae H, et al. Effects of glucose level on early and long-term mortality after intracerebral haemorrhage: the Acute brain bleeding analysis study. Diabetologia. 2010;53(3):429–34.

    Article  PubMed  CAS  Google Scholar 

  44. Wu T, Putaala J, Sharma G et al. Persistent hyperglycemia is Associated with increased mortality after Intracerebral Hemorrhage. J Am Heart Association 2017; 6(8).

  45. Capes S, Hunt D, Malmberg K, Pathak P, Gerstein H. Stress hyperglycemia and prognosis of stroke in nondiabetic and diabetic patients: a systematic overview. Stroke. 2001;32(10):2426–32.

    Article  PubMed  CAS  Google Scholar 

Download references

Acknowledgements

None.

Funding

The project was supported by the Innovation Science Fund of Tangdu hospital, China (No. 2023BTDQN001).

Author information

Author notes

  1. Shengru Liang, Xiaoxi Tian and Fei Gao contribute equally to this work and are co-first authors.

  2. Yang Yang and Lihong Li contributed equally to this work and are co-corresponding authors.

Authors and Affiliations

  1. Department of Endocrinology, Tangdu Hospital, Fourth Military Medical University, 710038, Xi’an, China

    Shengru Liang

  2. Department of Emergency, Tangdu Hospital, Fourth Military Medical University, 710038, Xi’an, China

    Xiaoxi Tian, Qi Wang, Jianwei Li, Lihong Li & Yang Yang

  3. Department of Neurosurgery, Tangdu Hospital, Fourth Military Medical University, 710038, Xi’an, China

    Fei Gao & Minghao Man

Authors
  1. Shengru Liang
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Xiaoxi Tian
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Fei Gao
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Minghao Man
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Qi Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Jianwei Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Lihong Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Yang Yang
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

SL, XT and FG analyzed the data and wrote the paper. YY and MM collected the data and checked the integrity of the data. QW and JL checked the accuracy of the data analysis. YY and LL designed the study and revised the paper. All authors have made an intellectual contribution to the manuscript and approved the final submission.

Corresponding author

Correspondence to Yang Yang.

Ethics declarations

Ethics approval and consent to participate

The study was performed according to the guidelines of the Helsinki Declaration. As the MIMIC-IV database is publicly available and all data are de-identified to remove patients’ information, the requirement for individual patient consent is not indispensable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Liang, S., Tian, X., Gao, F. et al. Prognostic significance of the stress hyperglycemia ratio and admission blood glucose in diabetic and nondiabetic patients with spontaneous intracerebral hemorrhage. Diabetol Metab Syndr 16, 58 (2024). https://doi.org/10.1186/s13098-024-01293-0

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s13098-024-01293-0

Keywords

Diabetology & Metabolic Syndrome

ISSN: 1758-5996

Contact us