Diabetic and stress‐induced hyperglycemia in spontaneous intracerebral hemorrhage: A multicenter prospective cohort (CHEERY) study

Abstract Introduction Admission hyperglycemia is a common finding after spontaneous intracerebral hemorrhage (ICH) secondary to pre‐existing diabetes mellitus (DM) or stress‐induced hyperglycemia (SIH). Studies of the causal relationship between SIH and ICH outcomes are rare. Aim We aimed to identify whether SIH or pre‐existing DM was the cause of admission hyperglycemia associated with ICH outcomes. Methods Admission glycosylated hemoglobin (HbA1c), glucose levels, and comorbidity data from the prospective, multicenter cohort, Chinese Cerebral Hemorrhage: Mechanisms and Intervention Study (CHEERY), were collected and analyzed. According to different admission blood glucose and HbA1c levels, patients were divided into nondiabetic normoglycemia (NDN), diabetic normoglycemia (DN), diabetic hyperglycemia (DH), and SIH groups. Modified Poisson regression models were used to analyze ICH outcomes in the different groups. Results In total, 1372 patients were included: 388 patients with admission hyperglycemia, 239 with DH, and 149 with SIH. In patients with hyperglycemia, SIH was associated with a higher risk of pulmonary infection [risk ratios (RR): 1.477, 95% confidence interval (CI): 1.004–2.172], 30‐day (RR: 1.068, 95% CI: 1.009–1.130) and 90‐day mortality after ICH (RR: 1.060, 95% CI: 1.000–1.124). Conclusions Admission hyperglycemia is a common finding after ICH, and SIH is a sensitive predictor of the risk of pulmonary infection and all‐cause death after ICH.


| INTRODUC TI ON
Spontaneous intracerebral hemorrhage (ICH) is the second most common and fatal type of stroke, 1 with a 30-day mortality rate of up to 40%. 2 Hyperglycemia is always observed at admission when one suffers from ICH and is considered a predictor of poor outcomes in some studies, [3][4][5][6][7] which were challenged by other studies. [8][9][10] This discrepancy may be due to the causes of admission hyperglycemia, which could either be stress-induced hyperglycemia (SIH) or preexisting diabetes mellitus (DM), which were not well differentiated in those studies. SIH is a transient hyperglycemic condition caused by acute diseases and is usually restricted to patients without DM, 11 and is defined as admission blood glucose ≥7.8 mmol/L, 12 which has been found to be directly responsive to the severity and predictable to the poor outcomes of ICH. [13][14][15][16] However, there are still some shortcomings in these well-designed reports: (1) up to 1/3 of patients with occult diabetes may clutter the results 17 as some previous studies defined SIH in non-DM patients only by disease history and not by measuring glycosylated hemoglobin A1c (HbA1c), 15,16,18 a reliable measure of the mean glucose concentration over the previous 3-to 4-month time period 19,20 ; (2) the few studies pertaining to SIH in patients with ICH were all relatively small populations (the largest included 328 patients) 9,13,14,16 ; and (3) some were retrospective studies. 14 Therefore, we sought to determine whether SIH or pre-existing DM was associated with poor outcomes of ICH in a large, multicentric, prospective cohort of patients with ICH having a long follow-up.

| Participants and design
We analyzed data from the Chinese Cerebral Hemorrhage: Mechanisms and Intervention study (CHEERY) ( (1) hemorrhages secondary to trauma, primary subarachnoid hemorrhage, hemorrhagic conversion from ischemic stroke, and thrombolysis; (2) lack of data on admission glucose or HbA1c levels; and (3) unavailability of imaging and baseline information.

| Data collection and follow-up
Relevant information was collected through the electronic medical record system: age, sex, disease history, admission systolic blood pressure (SBP), admission blood glucose, HbA1c level, time from symptom onset to admission, and surgical treatment. Admission blood glucose and HbA1c were measured after an overnight 8-h fast. Baseline neurological deficits were assessed using the Glasgow Coma Scale (GCS). 21 Hematoma localization and intraventricular hemorrhage (IVH) were recorded according to the first head CT data after admission, and hematoma volume was calculated using the ABC/2 formula. 22 23 and poor outcome was defined as mRS score of 3-6. 26-28

| Definition of subgroup
According to the latest consensus from the American Association of Clinical Endocrinologists and American Diabetes Association, 12 SIH was defined as having no DM history, HbA1c <6.5%, and admission blood glucose ≥7.8 mmol/L. If the admission blood glucose <7.8 mmol/L, it was defined as nondiabetic normoglycemia (NDN).
Diabetic hyperglycemia (DH) was defined as having a DM history or HbA1c ≥6.5%, and admission blood glucose ≥7.8 mmol/L; and if admission blood glucose <7.8 mmol/L, it was defined as diabetic normoglycemia (DN).

| Statistical analysis
SPSS statistical software (version 26.0, SPSS Corporation, Chicago) was used to analyze the data, and the statistical significance was set at p < 0.05. Categorical variables are expressed as percentages, and the χ 2 test or Fisher's exact test was used to compare the differences between groups. The Kolmogorov-Smirnov test (KS test) for normality was used to assess the data distribution of continuous variables. Normally distributed variables are expressed as mean ± standard deviation (SD), and two groups were compared using Student's t-test. Non-normally distributed variables were expressed as median and interquartile ranges (first and third quartiles), and the Mann-Whitney U test or Kruskal-Wallis H test was used to compare differences between groups. In univariate analysis, variables reaching p < 0.05 were considered to have significant differences. Finally, age, sex, history of hypertension, time from symptom onset to admission, infratentorial hemorrhage, hematoma volume, GCS, IVH, and surgical treatment were included in the regression model in the multivariate analysis. Modified Poisson regression models were used to calculate risk ratios (RRs) 29 and the associated 95% confidence intervals (CIs) for the association between different groups and the outcomes of interest, and figures were drawn using GraphPad Prism 9.0.

| Characteristics and outcomes of patients with different admission blood glucose levels
The baseline characteristics and outcomes of ICH patients with hyperglycemia or normoglycemia are compared in Table 1. The mean age of the included patients was 62.3 ± 11.6 years, 904 (65.9%) patients were male, and 281 (20.5%) patients had a history of DM.
The average age, sex ratio, admission SBP, hematoma volume, and length of hospitalization were not significantly different between the hyperglycemic and normoglycemic groups. In the hyperglycemia group, the median of admission blood glucose was 9.5 (8.5-11.5) mmol/L, and the mean HbA1c was 6.4%, both of which were higher than those in the normoglycemic group (p < 0.001

| Multivariate analysis for association between hyperglycemia and patient outcomes
The multivariate analysis of the association between hyperglycemia and ICH outcomes is shown in

| Characteristics and outcomes of ICH patients with different glucose classifications
Hyperglycemia at admission may be caused by SIH or DM. According to the HbA1c level, glucose level, and DM history, patients were classified into four groups: NDN, DN, DH, and SIH, as previously mentioned. The characteristics and outcomes of the four classifications are presented in Table 3. Overall, no significant differences in sex (p = 0.446) and age (p = 0.353) among the four groups were found. Compared with the other groups, patients with DH had a significantly larger hematoma volume (p < 0.001) and a higher incidence TA B L E 1 Baseline characteristics and outcomes of ICH patients with different admission blood glucose levels

| Multivariate analysis for outcomes of patients with ICH and different glucose classifications
The results of the multivariate regression models (

| DISCUSS ION
In this study, we found that hyperglycemia was associated with poor prognosis and an increased risk of death after ICH onset. After multivariate regression analysis, admission hyperglycemia was an independent risk factor for pulmonary infection, poor 30-and 90day prognosis, and 30-day mortality but did not increase the risk of 90-day mortality. 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.
Previous studies have shown that ICH is often accompanied by hyperglycemia, and the association between admission TA B L E 3 Baseline characteristics and outcomes of patients with different glucose classifications hyperglycemia and the risk of death and adverse outcomes of ICH has been concerning. [3][4][5][6][7][8][9][10] Hyperglycemia leads to peripheral nerve injury, 30 and hematoma perihematomal cell death 31 and decreased autophagy, 32 increasing the production of superoxide caused by tissue plasminogen activator, 33 and increasing the plasma kallikrein to promote the expansion of hematoma. 34  was significantly associated with the risk of SAP in patients without diabetes. 52 Additionally, hyperglycemia reportedly leads to the excessive release of inflammatory factors, such as tumor necrosis factorα (TNFα), interleukin-1 (IL-1) and interleukin-6 (IL-6), 53,54 which were significant contributors to pulmonary infection. [55][56][57][58] Simultaneously, increased proinflammatory factors and immunosuppression caused by stroke promote and accelerate the occurrence of pulmonary infection. 59 This study had several limitations. Although the sample size was large, there are few patients with DH and SIH, and only Chinese patients were included. Additionally, due to medical insurance policies and costs, the proportion of patients for whom HbA1c was measured was low, which was the most important reason restricting the inclusion of patients. Furthermore, some other variables known to be associated with poor outcomes of ICH, such as hematoma expansion, were not analyzed in this study due to unavailability of data, which should be further studied in the future.
In conclusion, admission hyperglycemia is common in ICH patients and is associated with poor outcomes, of which SIH may be prioritized over DH to predict the risk of pulmonary infection and 30-and 90-day death due to ICH.

AUTH O R CO NTR I B UTI O N S
Shaoli Chen, Yan Wan, and Hongxiu Guo conducted data analysis and wrote the manuscript. Gang Li, Quanwei He, and Bo Hu designed the study and wrote the manuscript. All authors helped with the data collection and literature searches. All the authors have approved this version of the manuscript for publication.

Dr. Bo Hu is an editorial board member of CNS Neuroscience and
Therapeutics and is a co-author of this article. To minimize bias, they were excluded from all editorial decision-making related to the acceptance of this article for publication.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data are available on request from the authors.