Impact of Stress‐Induced Hyperglycemia on In‐Hospital Medical Complications in Patients With Acute Stroke: From a Large‐Scale Nationwide Longitudinal Registry

ABSTRACT

Aims

This study aimed to explore the relationship between stress‐induced hyperglycemia (SIH) and in‐hospital medical complications in patients with acute stroke.

Methods

We enrolled 865,765 patients with acute stroke from the Chinese Stroke Center Alliance cohort. The stress hyperglycemia ratio (SHR) was defined as the ratio of fasting blood glucose to glycosylated hemoglobin. The primary outcomes were in‐hospital medical complications, including urinary tract infection (UTI), pneumonia, deep venous thrombosis (DVT), pulmonary embolism (PE), gastrointestinal bleeding (GIB), and pressure sores. Logistic regression analyses were performed to assess the association between the SHR and in‐hospital medical complications.

Results

In total, 104,873 (12.11%) patients with acute stroke experienced at least one in‐hospital medical complication. The most frequently observed complications were pneumonia (9.84%), UTI (1.29%), and GIB (1.06%). Both univariate and multivariate logistic regression analyses showed that a higher SHR was associated with a higher prevalence of pneumonia, UTI, GIB, and pressure sores (P for trend < 0.01). After adjusting for potential confounders, a higher SHR was associated with a reduced risk of PE (P for trend < 0.05), whereas no significant association was found between SHR and DVT. In the subgroup analysis, a significant association between elevated SHR and the occurrence of in‐hospital complications was further confirmed.

Conclusions

SIH is significantly associated with in‐hospital medical complications, particularly pneumonia, UTI, GIB, and pressure sores, in patients with acute stroke.

1. Introduction

Stroke is the second leading cause of mortality worldwide and remains a global public health challenge. The economic burden of stroke has increased from 1990 to 2019, particularly in low‐ and lower‐middle‐income countries [1]. Medical complications are common in patients with stroke, and accumulating evidence indicates that post‐stroke complications during hospitalization are related to prolonged length of stay, increased mortality, and poor functional outcomes [2, 3, 4, 5]. Data from the United States have shown that infection is the most common complication in patients with acute ischemic stroke (AIS). Among these, urinary tract infection (UTI) and pneumonia accounted for 11.8% and 3.2%, respectively. The incidence rates of gastrointestinal bleeding (GIB) (1.1%), deep vein thrombosis (DVT) (1.2%), and pulmonary embolism (PE) (0.5%) were comparatively low. However, a 40% increase in the risk of DVT was observed, and the prevalence of PE tripled from 2007 to 2019 [6]. The incidence rates of UTI, pneumonia, DVT, and PE following intracerebral hemorrhage (ICH) were 14.8%, 7.8%, 2.7%, and 0.7%, respectively [7]. In all hospitalized adult patients, the worldwide occurrence of pressure injuries is 12.8% [8]. Therefore, it is important to identify early predictive factors for complications, implement preventive measures, and provide appropriate treatment for complications during hospitalization.

Hyperglycemia upon admission is a common phenomenon in patients with acute stroke and may be caused by diabetes mellitus (DM) or stress‐induced hyperglycemia (SIH). SIH is transient hyperglycemia secondary to immune‐neurohormonal derangements caused by severe illnesses, including stroke [9]. Previous studies have shown that hyperglycemia on admission is associated with infections [10], unfavorable functional outcomes, and mortality [11, 12, 13] in patients with stroke. However, these studies were conducted in selected patient populations, focusing on patients with AIS, and limited research currently exists on the relationship between SIH and in‐hospital complications.

However, SIH currently lacks a standardized definition. When evaluating the relationship between SIH and critical illness outcomes, it is crucial to consider the background blood glucose levels, particularly in patients with diabetes. Glycosylated hemoglobin (HbA1c) is a well‐validated measure that represents the average glucose concentration over a period of approximately 2 – 3 months [14]. To better identify and quantify SIH, Roberts et al. first proposed the ratio of stress hyperglycemia (SHR), defined as the ratio of blood glucose to HbA1c [15]. Many studies have suggested that SHR could serve as reliable indicators of SIH with the inclusion of background glucose levels [16, 17].

Therefore, the aim of this study was to investigate the impact of SIH on in‐hospital medical complications (UTI, pneumonia, DVT, PE, GIB, and pressure sores) among patients with stroke (including both ischemic and hemorrhagic stroke) and its potential for guiding blood glucose management.

2. Methods

2.1. Study Cohort and Participants

This study used data from the Chinese Stroke Center Alliance (CSCA) [18]. This national multicenter cohort enrolled 1,006,798 hospitalized patients with stroke and TIA from 1476 hospitals. The CSCA study was initiated in August 2015 and completed in July 2019, similar to the American Heart Association's Get With the Guidelines‐Stroke program in the United States [19]. The inclusion criteria for CSCA study were (1) aged $\ge$ 18 years old, (2) primary diagnosis of acute stroke, including AIS, ICH, and subarachnoid hemorrhage (SAH) confirmed by brain computed tomography or magnetic resonance imaging and (3) admitted within 7 days of stroke onset, either directly or via the emergency department. Patients were excluded from our study if they had incomplete information on in‐hospital complications, age, sex, smoking, alcohol status, and important medical history, as well as patients without HbA1c, fasting blood glucose (FBG), and baseline National Institute of Health Stroke Scale (NIHSS) score data (Figure 1). The study was conducted in accordance with the Helsinki Declaration and approved by from the Ethics Committees of Beijing Tiantan Hospital and local hospitals.

FIGURE 1.

Study flowchart diagram.

2.2. Baseline Characteristics and Data Collection

In the CSCA, baseline information of the enrolled patients was entered directly by trained registrants using a web‐based patient data collection and management tool. Baseline clinical characteristics were extracted from the database, including demographic characteristics (age, sex, smoking, and alcohol status), medical history (hypertension, diabetes, dyslipidemia, atrial fibrillation, and stroke), prior medication use (antihypertensive, antiplatelet, and anticoagulant use), laboratory data, and clinical data (baseline NIHSS score, baseline modified Rankin Scale [mRS]). Diabetes referred to having a self‐reported history or admission HbA1c ≥ 6.5%. The NIHSS is a comprehensive clinical assessment tool comprising 11 domains designed to objectively quantify stroke severity. Because of its reliability and utility, the NIHSS has been widely adopted in clinical practice [20]. The mRS is a standardized neurological assessment tool used to evaluate the degree of disability and functional independence of patients. It employs a six‐point ordinal scale ranging from 0 – 5, where a score of 0 signifies no symptoms and a score of 5 indicates severe disability necessitating constant care and assistance [21].

2.3. Evaluation of Stress Hyperglycemia

Blood samples were obtained from the antecubital vein after overnight fasting for at least 8 h within the first 24 h of admission and examined for FBG and HbA1c levels. SIH was evaluated using the SHR, which was calculated using the following formulas: SHR = FBG (mmol/l)/HbA1c (%).

2.4. Outcome Assessment

The primary outcomes of interest were in‐hospital medical complications, including UTI, pneumonia, DVT, PE, GIB, and pressure sores. Complications were identified using the electronic medical records of the local treatment hospitals. The diagnosis of pneumonia was established based on the following criteria: presence of respiratory crackles, new purulent sputum, or positive sputum culture supported by typical findings on chest X‐ray. The diagnosis of UTI was established when a patient presented with clinical symptoms of UTI, in conjunction with a positive urine examination or culture. PE was clinically diagnosed based on radiographic evidence. The same principle applies to DVT. GIB was diagnosed based on clinical evidence such as fresh blood, coffee ground emesis, hematemesis, melena, or hematochezia, accompanied by laboratory or radiographic findings indicative of GIB. Pressure sores refer to any skin lesions or necrosis that occur because of pressure or minor trauma [18].

2.5. Statistical Analysis

Continuous variables are represented as mean ± standard deviation or median (interquartile range), while categorical variables are expressed as frequencies and percentages. Baseline characteristics were compared among patients stratified into different quartiles of SHR. Student's t‐test or the Mann–Whitney U test was used for comparisons. Categorical variables were compared using either the χ [2] test or Fisher's exact test. Univariate and multivariate logistic regression analyses were conducted to determine the associations between the SHR and in‐hospital medical complications in patients with stroke, which were expressed as unadjusted or adjusted odds ratios (ORs) with 95% confidence intervals (CIs). Multivariate logistic regression analyses were performed and adjusted for potential covariates including age, sex, NIHSS score at admission, alcohol consumption, current smoking status, history of hypertension, history of dyslipidemia, history of stroke, history of atrial fibrillation, DM, antidiabetic agents, antiplatelet agents, anticoagulation agents, antihypertension agents, and stroke classification. Additionally, subgroup analyses stratified by sex, age, diabetes, and stroke classification were performed and their interactions were explored. All statistical analyses were performed using the SAS statistical software (version 9.4 SAS Institute Inc.). Statistical tests were two‐tailed, and p < 0.05 was considered statistically significant.

3. Results

3.1. Baseline Characteristics

A total of 865,765 patients were included in this study. Patients were divided into four groups based on their SHR quartiles. Patients in the higher SHR quartile were more likely to have a history of hypertension (p < 0.001), higher baseline NIHSS score (p < 0.001), higher baseline mRS score (p < 0.001), and a higher proportion of hemorrhagic stroke. As the SHR increased, a gradual decrease in the proportion of male patients (p < 0.001), smoking status (p < 0.001), and the proportion of ischemic stroke (p < 0.001) was observed. There were also significant differences in age, alcohol consumption, history of stroke, history of DM, history of dyslipidemia, history of atrial fibrillation, and previous use of antihypertensive drugs, anticoagulants, antiplatelet drugs, and antidiabetic drugs among the different quartiles of SHR (all P values < 0.001) (Table 1).

TABLE 1.

Baseline characteristics of the patients by quartile of the stress hyperglycemia ratio.

Variables	Q1 (SHR ≤ 0.88)	Q2 (0.88 < SHR ≤ 1.00)	Q3 (1.00 < SHR ≤ 1.16)	Q4 (SHR > 1.16)	p
	n = 216, 502	n = 228, 641	n = 204, 211	n = 216, 411
Demographics characteristics, mean ± SD/n (%)
Age, years	66.7 ± 12.1	65.8 ± 12.2	65.6 ± 12.1	65.7 ± 12.0	< 0.001
Male	139,721 (64.5)	144,193 (63.1)	126,895 (62.1)	127,623 (59.0)	< 0.001
Alcohol consumption	50,260 (23.2)	54,543 (23.9)	49,697 (24.3)	50,059 (23.1)	< 0.001
Current smoking	56,450 (26.1)	56,824 (24.9)	48,845 (23.9)	46,914 (21.7)	< 0.001
Baseline clinical date, Median (IQR)
Baseline mRS	1.0 (1.0–2.0)	1.0 (1.0–2.0)	1.0 (1.0–2.0)	2.0 (1.0–3.0)	< 0.001
Baseline NIHSS	3.0 (1.0–6.0)	3.0 (1.0–6.0)	3.0 (2.0–6.0)	4.0 (2.0–8.0)	< 0.001
Medical history, n (%)
Diabetes mellitus	40,289 (18.6)	32,210 (14.1)	39,203 (19.2)	73,262 (33.9)	< 0.001
Hypertension	137,494 (63.5)	146,570 (64.1)	134,026 (65.6)	145,343 (67.2)	< 0.001
Dyslipidemia	15,224 (7.0)	15,439 (6.8)	14,647 (7.2)	18,194 (8.4)	< 0.001
Atrial fibrillation	10,933 (5.0)	10,663 (4.7)	9841 (4.8)	11,746 (5.4)	< 0.001
Stroke	148,136 (68.4)	158,824 (69.5)	139,842 (68.5)	144,740 (66.9)	< 0.001
Prior medication use, n (%)
Antiplatelet	43,595 (20.1)	43,526 (19.0)	38,935 (19.1)	41,889 (19.4)	< 0.001
Anticoagulation	7182 (3.3)	6866 (3.0)	6033 (3.0)	7027 (3.2)	< 0.001
Antihypertension	100,921 (46.6)	105,839 (46.3)	97,896 (47.9)	106,573 (49.2)	< 0.001
Antidiabetic	183,809 (84.9)	203,816 (89.1)	173,936 (85.2)	158,479 (73.2)	< 0.001
Laboratory indicators, mean ± SD
FBG, mmol/l	4.9 ± 1.3	5.7 ± 1.2	6.4 ± 1.7	9.2 ± 4.0	< 0.001
HbA1c, %	6.6 ± 1.9	6.0 ± 1.3	6.0 ± 1.5	6.3 ± 2.1	< 0.001
Stroke classification, n (%)
Ischemic stroke	204,185 (94.3)	212,712 (93.0)	186,905 (91.5)	191,571 (88.5)	< 0.001
Intracranial hemorrhage	10,594 (4.9)	13,895 (6.1)	15,116 (7.4)	20,994 (9.7)
Subarachnoid hemorrhage	825 (0.4)	1121 (0.5)	1318 (0.6)	2905 (1.3)
Not classified	898 (0.4)	913 (0.4)	872 (0.4)	941 (0.4)

Abbreviations: FBG, fasting blood glucose; HbA1c, glycosylated hemoglobin; mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale.

3.2. Association Between SIH and In‐Hospital Complications

A total of 12.11% of patients (n = 104,873) experienced at least one in‐hospital medical complication. Approximately 1.42% of the patients (n = 12,314) experienced more than two complications during hospitalization (Figure 2A). The most common complications were pneumonia (n = 85,224), UTI (n = 11,181), and GIB (n = 9309) (Figure 2B).

FIGURE 2.

The distribution of patients with multiple complications (A) and the number of patients with different in‐hospital medical complications (B). N represents the number of complications that patients experienced during hospitalization.

In both univariate and multivariate logistic regression analyses, the SHR was positively associated with in‐hospital complications, including pneumonia, GIB, pressure sores, and UTI (P for trend < 0.01). No significant association was observed between SHR and DVT in the univariate analysis (P for trend = 0.06). Although significant associations were observed in Model 1 and Model 2 (P for trend < 0.05), no significant association was found in Model 3 (P for trend = 0.05). Although univariate regression did not show a significant association between SHR and PE (P for trend > 0.05), a higher SHR was significantly associated with a lower incidence of PE in both Model 2 and Model 3 (P for trend < 0.05) (Table 2).

TABLE 2.

Association of stress‐induced hyperglycemia with in‐hospital medical complications.

Complication	n (%)	Crude	Model 1	Model 2	Model3
		OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)
Pneumonia
Q1	17,445 (8.1)	1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)
Q2	18,726 (8.2)	1.02 (1.00,1.04)	1.06 (1.04,1.08)	1.01 (0.99,1.03)	1.01 (0.99,1.03)
Q3	19,671 (9.6)	1.22 (1.19,1.24)	1.29 (1.26,1.32)	1.14 (1.11,1.16)	1.14 (1.11,1.17)
Q4	29,372 (13.6)	1.79 (1.76,1.83)	1.92 (1.88,1.95)	1.42 (1.38,1.45)	1.41 (1.38,1.44)
P trend		< 0.01	< 0.01	< 0.01	< 0.01
Gastrointestinal bleeding
Q1	1514 (0.7)	1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)
Q2	1756 (0.8)	1.10 (1.03,1.18)	1.13 (1.05,1.21)	1.08 (1.01,1.16)	1.08 (1.01,1.16)
Q3	2060 (1.0)	1.45 (1.35,1.55)	1.49 (1.40,1.60)	1.32 (1.24,1.42)	1.32 (1.23,1.41)
Q4	3851 (1.8)	2.57 (2.42,2.73)	2.66 (2.50,2.82)	1.85 (1.74,1.97)	1.85 (1.74,1.97)
P trend		< 0.01	< 0.01	< 0.01	< 0.01
Deep vein thrombosis
Q1	2368 (1.1)	1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)
Q2	2251 (1.0)	0.90 (0.85,0.95)	0.91 (0.86,0.97)	0.90 (0.85,0.95)	0.90 (0.85,0.96)
Q3	2041 (1.0)	0.91 (0.86,0.97)	0.93 (0.88,0.99)	0.87 (0.82,0.92)	0.87 (0.82,0.93)
Q4	2649 (1.2)	1.12 (1.06,1.19)	1.13 (1.07,1.20)	0.94 (0.88,0.99)	0.95 (0.89,1.00)
P trend		0.06	< 0.01	0.02	0.05
Pulmonary embolism
Q1	539 (0.2)	1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)
Q2	507 (0.2)	0.89 (0.79,1.01)	0.90 (0.80,1.01)	0.89 (0.79,1.00)	0.89 (0.79,1.00)
Q3	369 (0.2)	0.73 (0.64,0.83)	0.73 (0.64,0.84)	0.69 (0.60,0.79)	0.70 (0.61,0.80)
Q4	532 (0.2)	0.99 (0.88,1.11)	0.99 (0.88,1.12)	0.86 (0.76,0.97)	0.87 (0.77,0.98)
P trend		0.29	0.34	< 0.01	< 0.01
Bedsore
Q1	530 (0.2)	1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)
Q2	550 (0.2)	0.98 (0.87,1.11)	1.03 (0.91,1.16)	0.98 (0.87,1.11)	0.99 (0.87,1.12)
Q3	596 (0.3)	1.19 (1.06,1.34)	1.26 (1.12,1.42)	1.10 (0.98,1.24)	1.11 (0.98,1.25)
Q4	1057 (0.5)	2.00 (1.80,2.22)	2.09 (1.89,2.33)	1.39 (1.25,1.55)	1.40 (1.25,1.56)
P trend		< 0.01	< 0.01	< 0.01	< 0.01
Urinary tract infection
Q1	2392 (1.1)	1.00 (ref)	1.00 (ref)	1.00 (ref)	1.00 (ref)
Q2	2541 (1.1)	1.01 (0.95,1.06)	1.02 (0.96,1.07)	1.02 (0.96,1.08)	1.02 (0.97,1.08)
Q3	2669 (1.3)	1.19 (1.12,1.25)	1.19 (1.13,1.26)	1.12 (1.06,1.18)	1.12 (1.06,1.19)
Q4	3579 (1.7)	1.51 (1.43,1.59)	1.47 (1.40,1.55)	1.15 (1.09,1.21)	1.15 (1.09,1.22)
P trend		< 0.01	< 0.01	< 0.01	< 0.01

Note: Model 1: adjusted for age, sex; Model 2: adjusted for covariates from Model 1 and further adjusted for National Institutes of Health Stroke score at admission, alcohol consumption, current smoking status, history of hypertension, history of dyslipidemia, history of stroke, history of atrial fibrillation, diabetes mellitus, antidiabetic agents, and stroke classification. Model 3: adjusted for covariates from Model 2 and further adjusted for history of antiplatelet, anticoagulation, and antihypertension agents.

3.3. Subgroup Analysis

The significant association between higher SHR and the occurrence of in‐hospital complications was further validated in almost all subgroups (p < 0.05). We also observed that a significant interaction effect existed among subgroups of sex (interaction p < 0.01), age (interaction p < 0.01), diabetes (interaction p < 0.01), and type of stroke (interaction p < 0.01). Notably, the relationship between SHR and in‐hospital medical complications was more pronounced in female patients, those aged ≥ 65 years, patients without diabetes, or with SAH (Table 3).

TABLE 3.

Subgroup analyses of SHR for in‐hospital medical complications.

		OR (95% CI)	p for interaction
Sex
Male	Q1	reference	< 0.01
	Q2	0.99 (0.96,1.01)
	Q3	1.09 (1.06,1.12)
	Q4	1.35 (1.31,1.38)
Female	Q1	reference
	Q2	1.02 (0.98,1.05)
	Q3	1.13 (1.09,1.16)
	Q4	1.36 (1.32,1.40)
Age $\ge$ 65
Yes	Q1	reference	< 0.01
	Q2	1.01 (0.99,1.04)
	Q3	1.14 (1.11,1.17)
	Q4	1.38 (1.35,1.41)
No	Q1	reference
	Q2	0.97 (0.93,1.00)
	Q3	1.05 (1.01,1.09)
	Q4	1.29 (1.25,1.34)
Diabetes mellitus
Yes	Q1	reference	< 0.01
	Q2	0.97 (0.91,1.02)
	Q3	0.97 (0.92,1.02)
	Q4	1.24 (1.19,1.30)
No	Q1	reference
	Q2	1.01 (0.99,1.03)
	Q3	1.13 (1.11,1.16)
	Q4	1.38 (1.35,1.41)
Type of stroke
Ischemic stroke	Q1	reference	< 0.01
	Q2	1.01 (0.99,1.03)
	Q3	1.13 (1.11,1.16)
	Q4	1.38 (1.35,1.41)
Intracranial hemorrhage	Q1	reference
	Q2	0.99 (0.97,1.01)
	Q3	1.09 (1.07,1.11)
	Q4	1.32 (1.29,1.34)
Subarachnoid hemorrhage	Q1	reference
	Q2	1.16 (0.92,1.47)
	Q3	1.36 (1.09,1.70)
	Q4	1.70 (1.39,2.07)

Abbreviations: CI, confidence interval; OR, odds ratio. Subgroup analyses were adjusted for National Institutes of Health Stroke score at admission, alcohol consumption, current smoking status, history of hypertension, history of dyslipidemia, history of stroke, history of atrial fibrillation, antidiabetic agents, antiplatelet agents, anticoagulation agents, and antihypertension agents.

4. Discussion

The CSCA is a national multicenter registry that enrolls a large number of patients who experienced acute stroke throughout China. In this study, we found a significant association between SHR and pneumonia, UTI, GIB, and pressure sores; patients with a higher SHR were more likely to have these complications. After adjusting for potential confounders, SHR was inversely associated with the occurrence of PE but was not significantly associated with DVT. However, further research is required to explore this relationship.

Admission hyperglycemia has been reported to be associated with post‐stroke pneumonia and UTI in non‐diabetic patients with AIS [10]. A previous study showed that hyperglycemia was not associated with infectious events after ICH [22]. However, these studies did not differentiate between SIH and chronic diabetes, both of which could lead to hyperglycemia upon admission. A few studies have explored the relationship between SIH and infectious complications [23, 24, 25]. A meta‐analysis demonstrated a significant increase in infectious complications after acute stroke (pneumonia [OR 2.06, 95% CI 1.57–2.72, p < 0.00001] and UTI [OR 2.53, 95% CI 1.45–4.42, p = 0.001]) in patients with higher SHR, which is consistent with our findings [17]. However, this meta‐analysis included only two studies on hemorrhagic stroke, and the proportion of patients with hemorrhagic stroke accounted for only 0.4% of the total sample size. The association between hyperglycemia and post‐stroke infections may be attributed to an immunocompromised state and higher oxidative response [9, 26]. Stroke‐induced immunosuppression involves activation of the autonomic nervous system, which triggers the release of immune mediators, leading to long‐lasting systemic immunodepression [27]. Acute hyperglycemia may exacerbate the immune immunocompromised state, particularly by reducing neutrophils' activity [28, 29].

In our study, only 0.3% of the patients with acute stroke experienced a pressure sore event during hospitalization. This result may be attributed to the relatively low NIHSS score of the study population (with an average NIHSS score of 3 at admission). Another reason could be the increased emphasis on early care for patients with stroke by medical staff in Chinese hospitals. The most prevalent risk factors for pressure sores development include three primary domains: mobility or activity, perfusion (including diabetes), and skin or pressure status [30]. A study conducted in a Chinese population suggested that diabetes was a risk factor for pressure sores in patients with AIS; however, elevated blood glucose levels at admission were not associated with pressure sores [31], indicating that long‐term damage caused by abnormal blood glucose levels is closely linked to their occurrence. Nevertheless, our findings indicated that higher SHR is correlated with an increased risk of developing pressure sores. Compared to other biomarkers [27], hyperglycemia is an intervention indicator deserving attention in future studies.

Whether hyperglycemia is a risk factor for venous thromboembolism (VTE) after an acute stroke is unclear. Several in vivo studies have demonstrated that hyperglycemia leads to increased levels of crucial proteins involved in the coagulation pathway, such as fibrinogen, prothrombin 1 and 2, tissue factor, thrombin‐antithrombin complexes, plasminogen activator inhibitor‐1, tissue plasminogen activator, and complement C3 [32, 33]. Moreover, hyperglycemia‐induced insulin resistance exacerbated endothelial dysfunction [34], activated platelets [35], and impaired fibrinolysis [36]. A previous study suggested that higher blood glucose levels are associated with a high risk of asymptomatic DVT in patients with acute stroke [24]. In another study, fasting glucose levels were associated with a greater risk of DVT after acute stroke. Nevertheless, in the multivariate analysis, fasting glucose levels were not significantly associated with DVT [37]. It was also reported that hyperglycemia was not linked to DVT in patients with AIS treated with intravenous recombinant tissue plasminogen activator [25]. We observed a significant inverse association between SHR and the incidence of PE. However, no significant association was found between elevated SHR and the risk of DVT after multivariate adjustment in Model 3. Future multicenter, large‐sample clinical trials are required to further elucidate the association between SIH and VTE.

GIB was the third most common (1.06%) complication of acute stroke, while it was one of the least prevalent (1.1%) in‐hospital complications in American patients with AIS between 2007 and 2019 [6]. In addition to pharmacological factors such as acetylsalicylic acid, several risk factors have been reported to be associated with the incidence of GIB, including age, sex, history of gastrointestinal diseases such as peptic ulcers, sepsis, shock [38], and adverse reactions to enteral tubes inserted [39]. In our study, the SHR was found to be an independent risk factor for GIB in patients with acute stroke. SIH is mediated by inflammatory responses and neuroendocrine derangements through pathways involving elevated levels of catecholamines, growth hormones, cortisol, and cytokines [9]. Stress ulcers may arise from the hyperactivity of the sympathetic nervous system, which triggers excessive catecholamine release and heightened secretion of gastric acid and pepsin, ultimately resulting in mucosal ischemia and damage to the gastrointestinal mucosa [40]. There may be shared pathways between SIH and stress ulcers; however, their reciprocal influence warrants further investigation.

Although hyperglycemia is associated with poor outcomes after stroke, intensive glycemic control is not recommended during the acute phase. A meta‐analysis that incorporated results from the SHINE trial (SHINE) indicated that tight glycemic control after AIS did not show any significant improvements in mortality, independence, or mRS score [41]. However, most studies have primarily focused on the relationship among intensive glycemic control, poor functional outcomes, and death after stroke, with only one trial encompassing post‐stroke complications. In the United Kingdom Glucose Insulin in Stroke Trial, no significant differences were observed in the incidence of chest or urinary infections after treatment with glucose–potassium–insulin compared with saline [42]. This study has faced criticism owing to its inclusion of a heterogeneous population, slow recruitment rate, delayed initiation of treatment, and inefficient glucose control. Consequently, there is insufficient and reliable evidence to determine whether intensive glycemic control can effectively improve in‐hospital complications. Several studies have suggested that hyperglycemia increases the risk of poor outcomes after stroke; however, no evidence has been found to support the notion that complications account for or mediate this relationship [10, 23, 25]. Therefore, further investigation is needed to ascertain whether intensive glycemic control is effective in reducing in‐hospital complications and its association with poor post‐stroke outcomes.

SIH may be associated with different clinical outcomes in patients with and without diabetes [17], which can be attributed to pathological changes in response to chronic hyperglycemia [43]. The response of patients with DM to fluctuations in glucose concentrations is diminished; however, the observed vascular complications may also be influenced by additional metabolic abnormalities [44, 45]. As some previous studies defined SIH only in patients without DM [46, 47, 48], they failed to distinguish SIH from newly diagnosed or unknown diabetes and did not consider background glucose levels [49]. Our subgroup analysis revealed a more pronounced association between SIH and complications in the patients without DM. Timely identification of patients with hyperglycemia requiring intervention, particularly in patients without DM, is highly significant in future studies.

The role of SIH may vary among the stroke types. The incidence of SIH at admission was observed in up to 24% of patients with AIS [5], whereas SIH occurred in approximately 50% of patients with SAH [42, 50] and 30% of patients with ICH [51]. Our study also revealed a gradual increase in the proportion of hemorrhagic stroke and a corresponding decrease in the proportion of ischemic stroke in the higher‐SHR group. The relationship between SIH and the development of hospital complications was particularly significant in patients with hemorrhagic stroke, as indicated by the subgroup analysis. However, different types of stroke have diverse complication rates. For instance, ICH may carry a higher risk of pneumonia and UTI compared with AIS [6, 7]. Investigating the impact of SIH on specific complications in different stroke types may provide valuable insights for identifying targeted and tailored treatment options in the future.

Our study has some limitations. First, the study had a retrospective design based on an observational cohort that inherently constrained the capacity to establish cause‐effect relationships. Second, the CSCA, as a quality improvement program for stroke treatment in the neurology department, targets the Chinese population and might have excluded patients in the neurosurgery or intensive care units, thereby introducing an inherent selection bias into the study, as most of the enrolled patients with stroke had relatively mild strokes. Therefore, this study may not be representative of the entire population with acute stroke. Third, the study did not assess the dynamic glycemic fluctuations and blood glucose management during hospitalization because of the absence of relevant data in the database. Fourth, in‐hospital complications were diagnosed and documented by local physicians, leading to inevitable variations in definitions and identifications. Nevertheless, the approach is practical and acceptable for large registries.

5. Conclusions

SIH is a significant and independent risk factor for pneumonia, UTI, GIB, and pressure sores in patients with acute stroke. Timely identification of SIH after a stroke is crucial for making informed decisions and formulating treatment strategies.

Author Contributions

Xintong Song: Conceptualization, Methodology, Validation, Formal analysis, Investigation, Writing, Original Draft, Writing, Review and Editing. Yi Ju: Investigation, Writing, Review and Editing, Funding acquisition. Hongqiu Gu: Methodology, Formal analysis, Investigation. Zhikai Zhu: Methodology, Formal analysis, Investigation. Zixiao Li: Data curation, Writing, Review and Editing. Qian Zhang: Conceptualization, Methodology, Validation, Data curation, Writing, Original Draft, Writing, Review and Editing, Supervision, Funding acquisition. Xingquan Zhao: Data curation, Writing, Review and Editing, Funding acquisition.

Ethics Statement

The study was conducted in accordance with the Helsinki Declaration and obtained approval from the Ethics Committee of Beijing Tiantan Hospital (KY2018‐061‐02) and local hospitals.

Conflicts of Interest

The authors declare no conflicts of interest.

Funding Statement

This work was funded by Noncommunicable Chronic Diseases‐National Science and Technology Major Project grants 2024ZD0522200 and 2024ZD0522203; National Key Research and Development Program of China grants 2022YFC3501100 and 2022YFC3501102; National Health Commission of the People’s Republic of China grant W2024SNKT23; National Natural Science Foundation of China grant GrantNo.82371302; Beijing Scholar grant 097.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.