Introduction
With high rates of disability and mortality, stroke is the primary common cause of adult disability and mortality in China [
1,
2]. Given the current large demographic changes in China and the changing prevalence of vascular risk factors, it is expected that the mortality rate from stroke would rise sharply in the coming decades [
3,
4]. The mortality rate of stroke patients in hospitals is high, and the length of stay is increasing. Therefore, it is crucial to carry out effective risk stratification for stroke patients in the hospital, establish optimal management strategies, and achieve accurate treatment [
5,
6].
Stress hyperglycemia affects a significant proportion of stroke patients [
7]. A rise in blood glucose levels has been noted in the acute phase of stroke. Stress hormones are released following activation of the sympathetic nervous system and the hypothalamus-pituitary-adrenal axis, causing this rise [
8]. Previous research suggests that persistent hyperglycemia is associated with more adverse outcomes in stroke incidents [
9]. Studies have shown that glycosylated hemoglobin (HbA1c) level was an independent predictor of worse functional outcome in patients with acute anterior circulation ischemic stroke (IS) [
10,
11]. Persistent hyperglycemia has been linked to worse outcomes from strokes, according to prior study [
9]. On the other hand, acute stress reactivity or persistent hyperglycemia may be the cause of increased glucose levels at the time of hospital admission [
12]. To address this issue, a novel marker that more precisely reflects acute hyperglycemia has been created: the stress hyperglycemia ratio (SHR). The chronic glycemic value, which is determined using glycosylated hemoglobin (HbA1c), and the acute admission blood glucose (ABG) are the basis for estimating the SHR [
13,
14]. Research indicates a significant independent correlation between SHR and adverse long-term outcomes in non-obstructive coronary artery patients [
14]. Research has shed light on the influence of SHR on initial neurological impairments and the future outcomes for patients with acute IS [
15]. Further research indicates that SHR serves as a reliable predictor of early hematoma growth and adverse outcomes in patients experiencing intracerebral hemorrhage [
16]. SHR is suggested as a promising indicator for predicting unfavorable outcomes in critically ill patients by assessing the extent of stress-induced hyperglycemia in relation to the severity of illness [
17]. However, the prognosis of patients with SHR and all types of stroke remains unclear. In addition, previous studies have limited attention given to the association between SHR and length of stay. Burdening society by prolonging hospitalization, which further increases the cost of healthcare [
18]. However, no study has yet explored the possible connection between SHR and length of stay. Thus, we performed a retrospective analysis to look into the relationship between SHR and in-hospital mortality as well as the length of stay in patients with stroke.
Materials and methods
Study population
Between January 1, 2015, and December 31, 2022, 10,466 acute stroke patients in total were included in the research. The following were the study’s inclusion criteria: (1) age ≥ 18 years; (2) acute stroke [including acute IS and hemorrhagic stroke (HS)] verified by radiological examination [cranial computed tomography (CT) or magnetic resonance imaging (MRI)]. The following were the exclusion criteria: (1) incomplete data pertaining to age, sex, body mass index (BMI), and comorbidities; (2) patients with conditions that affect HbA1c levels, such as renal failure (serum creatinine concentration > 180 mmol/L) and anemia (hemoglobin < 100 g/L); (3) lack of baseline data for HbA1c or admission glucose levels, or measurements not taken within 24 h of admission; (4) ABG < 3.9 mmol/L; (5) patients who were discharged or died within 48 h of admission. Ultimately, 4018 participants were included in the analysis (Figure
S1). All patients (or their legal relatives) who participated in the study gave written informed permission, and the study was approved by the hospital’s ethics committee. We followed the STROBE statement in reporting this study [
19].
Data collection and definition
Baseline data, including clinical information, lifestyle variables, test findings, medical history, and medication history, were taken from the electronic medical record. Clinical data included age, sex, height, weight, BMI, systolic, diastolic, and heart rate at admission. Both smoking and alcohol drinking were classified as current or non-current.
Laboratory indicators include white blood cells (WBC), red blood cells (RBC), hemoglobin (HB), platelets (PLT), alanine aminotransferase (ALT), aspartate aminotransferase (AST), albumin/globulin ratio (A/G), serum creatinine (Scr), uric acid (UA), serum potassium, serum sodium, total cholesterol (TC), triglycerides (TG), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), and HbA1c. The ABG was defined as the first available plasma glucose measurement within a day after admission. Relative hyperglycemia was determined by the SHR, calculated as the admission blood glucose (ABG) divided by the estimated average glucose [
13]. The SHR is expressed by the formula: SHR = ABG (mmol/L) / [1.59 × HbA1c (%) − 2.59] [
13]. The above blood biochemical indicators were detected by Automatic Analyzer (7600-010, Hitachi, Tokyo, Japan) according to the manufactures instruction.
The definition of hypertension was a self-reported history of medical diagnosis and antihypertensive drug usage. Dyslipidemia was defined by self-report dyslipidemia and/or treatment with a lipid-lowering drug. In case the participant disclosed a history of diabetes mellitus or used anti-hyperglycemic medication, diabetes mellitus was deemed to be present. The diagnosis of coronary artery disease was based on the patient’s self-reported history of myocardial infarction, percutaneous coronary intervention, or coronary bypass grafting. An ischemic or hemorrhagic stroke history was considered a prior stroke.
Outcomes
The outcome measures for this study included the length of stay and in-hospital mortality. Whatever the cause of death, every death that happened while the patient was in the hospital was considered to be an instance of in-hospital mortality. The time from admission to discharge was considered the length of stay. The medical records of stroke patients were identified by trained staff using the International Classification of Diseases, Tenth Revision (ICD-10) codes. The diagnosis and classification of acute stroke were determined based on WHO criteria, in combination with confirmation through brain CT or MRI [
20,
21].
Statistical analysis
The χ2 test was used for categorical variables, and for continuous variables, either ANOVA or the Kruskal-Wallis test was employed.
We calculated the variance inflation factor (VIF) for all covariates (Table
S1). Continuous outcomes were analyzed using multivariable linear regression, and binary outcomes were analyzed using multivariable logistic regression. Tests for trend were calculated using the tertile number as a continuous variable. Spearman’s correlations were employed to assess the association between the risk factors linked to in-hospital mortality and length of stay and SHR levels as a continuous variable. Restricted cubic spline (RCS) models were fitted for both linear and logistic regression models. To assess potential interactions, stratified analyses were carried out. Sensitivity analyses were employed to test the robustness of the results. Furthermore, the receiver operating characteristics (ROC) curve was plotted to evaluate the predictive utility of SHR for in-hospital mortality compared to ABG and HbA1c. Statistical analysis was performed using R 4.1.1 software. We considered
p < 0.05 (two sided) as significant.
Discussion
This is the first study to examine the relationship between SHR and in-hospital mortality and length of stay in stroke patients. The current study has two major findings. First, among stroke patients, those with high SHR had higher in-hospital mortality and length of stay. Second, SHR is independently associated with in-hospital mortality and length of stay in ischemic and hemorrhagic stroke patients.
The study determined that the SHR is a more effective quantitative indicator for stress hyperglycemia than blood glucose levels when assessing critical illness outcomes
11. Previous research has demonstrated a negative correlation between stress hyperglycemia and poor prognosis in patients with acute myocardial infarction [
22‐
24]. P. A. O’Neill et al. have previously determined the stress hormone and blood glucose response after acute stroke and found that irritability and blood glucose increases in non-diabetic patients are signs of a poor prognosis after stroke [
25]. A recent meta-analysis demonstrated that higher SHR levels significantly increase the rates of poor prognosis, mortality, neurologic deficits, hypertension, and infectious complications, regardless of diabetes or whether patients underwent intravenous thrombolysis or mechanical thrombectomy [
26]. In a prospective study, Konstantinos Tziomalos et al. discovered that stress hyperglycemia is linked to more severe acute IS and that such patients had more adverse functional outcomes and higher in-hospital mortality at discharge
13. These findings suggest that SHR could be a significant factor in patient prioritizing and a valid biomarker for predicting in-hospital mortality in stroke populations. Acute hyperglycemia following a stroke might be a risk factor for mortality since poor glycemic control seems to be linked to post-stroke dysfunction. Cao et al. found that hyperglycemia can enhance brain damage through multiple mechanisms, prolong length of hospital stay, increase recurrence and mortality rates, and negatively impact prognosis [
27]. Other studies have also shown that the longer critically ill patients stay in the emergency department, the higher their morbidity and mortality rates [
28]. Therefore, early identification of novel risk factors during the early stages of cerebral infarction and the implementation of hypoglycemic treatments are critical to reducing the harmful consequences of hyperglycemia.
Several potential mechanisms may account for the observed associations. First, stress hyperglycemia might be involved in the production of molecules that are inflammatory and vasoconstrictive, aggravating the oxidative stress that follows, causing mitochondrial malfunction, and results in endothelial dysfunction [
29‐
31]. Second, hyperglycemia also activates protein kinase C and NADPH oxidase, increasing reactive oxygen species (ROS) levels and reducing nitric oxide synthase, which cause decreased reperfusion and aggravate more neurons to be injured [
32]. Third, patients with acute IS often suffer from lipid metabolism disorders. Under the high blood sugar situation, macrophages swallow the glycated LDL, which will be transformed into foam cells, adhering to the blood vessel wall, thus contributing to the overlapping interaction of lipotoxicity, glucotoxicity, and inflammation, and causing poor clinical outcomes [
33]. Finally, hyperglycemia may be directly neurotoxic to the ischemic penumbra, causing more neurons to be damaged. This might then cause hypoperfused at-risk tissue to infarct, thus resulting in poorer stroke outcome [
34‐
36]. This would also indicate that SHR was a target for treatment in an attempt to alleviate the dismal prognosis associated with stroke.
Strengths and limitations
Our study has several strengths, including its large sample size, consistent data collection methods, and comprehensive information on potential confounders. The study, however, is not without limitations. First, as this is a retrospective, single-center investigation, further patient cohorts need be used to confirm our findings. Second, a direct causal relationship cannot be inferred due to the study’s design. Third, the study failed to gather data on the hypoglycemic treatment administered to the participants, thereby preventing the assessment of any additional correlations between SHR and the hypoglycemic treatment. Fourth, since SHR was only measured at one moment in time, we were unable to account for variations in SHR over time, which could possibly affect in-hospital mortality. Fifth, we did not collect data on enteral nutrition, corticosteroids, or other interventions that might affect in-hospital mortality. Sixth, due to the absence of baseline scores from the National Institutes of Health Stroke Scale (NIHSS) and the Glasgow Coma Scale (GCS) in our dataset, our analysis could not ascertain the influence of SHR on NIHSS and GCS scores.
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.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.