Stress hyperglycemia may have higher risk of stroke recurrence than previously diagnosed diabetes mellitus

We aim to evaluate the risk of stroke recurrence among non-diabetes mellitus (non-DM), previously diagnosed diabetes mellitus (PDDM), newly diagnosed diabetes mellitus-related hyperglycemia (NDDM-RH) and stress hyperglycemia after minor stroke or TIA. Totally, 3026 patients with baseline fasting glucose and glycated albumin from the CHANCE trial (Clopidogrel in High-Risk Patients with Acute Nondisabling Cerebrovascular Events) were included. Patients were classified as non-DM, PDDM, NDDM-RH and stress hyperglycemia according to the status of glucose metabolism. The primary outcome was stroke recurrence during 90-day follow up. Cox regression was performed to estimate the relationship between the status of glucose metabolism and risk of 90-day stroke recurrence. Compared with PDDM, NDDM-RH had a similar risk of 90-day stroke recurrence (hazard ratios [HR]1.39, 95% confidence intervals [CI] 0.94-2.04), while stress hyperglycemia had approximately a 5.3-fold increased risk of 90-day stroke recurrence after adjusted for confounding covariates (HR 5.32, 95% CI 3.43-8.26). Parallel results were found for 90-day recurrent ischemic stroke and composite events. Compared with PDDM in minor stroke or TIA, a parallel risk of 90-day stroke recurrence were observed for NDDM-RH, while stress hyperglycemia might relate to higher risk of 90-day stroke recurrence.


INTRODUCTION
Minor stroke and transient ischemic attack (TIA) comprises 65% among acute ischemic cerebrovascular diseases. [1] Patients with TIA or minor stroke are related to higher stroke risk after symptom onset at the early period. [2,3] Although intensified dual antiplatelet therapy, a small portion of patients with minor stroke or TIA still have recurrent stroke. [4][5][6] Diabetes mellitus (DM) has been demonstrated as a independent predictor of stroke recurrence after index events of ischemic stroke or TIA. [7][8][9][10] Patients with DM have a worse vascular prognosis than nondiabetic patients. [11] Furthermore, DM have been widely applied for predicting outcomes after acute ischemic stroke or TIA. [12,13] Consequently, the presence of pre-existing DM have received much attention in stroke patients. Newly diagnosed diabetes mellitus-related hyperglycemia (NDDM-RH) can predict one-year stroke recurrence, death, and poor functional outcome compared with non-diabetes mellitus (non-DM) in ischemic stroke. [14] Stress hyperglycemia has been identified as a predictor of stroke recurrence and poor outcome after TIA or ischemic stroke. [15,16] A previous study has shown that NDDM was associated with higher risk of death [17] and more likely to have poorer functional outcome and more severe strokes than patients with previously diagnosed diabetes mellitus (PDDM). [18] However, very few studies have compared the risk of recurrent stroke among non-DM, PDDM, NDDM-RH and stress hyperglycemia after minor stroke or TIA. Previous studies have shown that patients with poor glucose control or diabetes mellitus were less sensitive to aspirin [19,20] and the interaction of glucose metabolism status according to therapy of aspirin only or combination of clopidogrel and aspirin is uncertain.
We aim to investigate the associations of non-DM, PDDM, NDDM-RH and stress hyperglycemia with outcomes in patients with minor stroke or TIA from the Clopidogrel in High-risk patients with Acute Nondisabling Cerebrovascular Events (CHANCE) trial. In addition, we aim to evaluate the interaction effect of different glucose metabolism status by treatment of aspirin only or combination of clopidogrel and aspirin after minor stroke or TIA.

Baseline characteristics
Among the 3044 consecutive patients enrolled in the prespecified biomarker substudy of CHANCE, 3026 (99.4%) patients with available fasting blood glucose and glycated albumin (GA) were included in this subanalysis. The median age of the total patients was 62.2 years and 66.5% of them were men. The median fasting blood glucose level was 5.5 mmol/L (interquartile, 4.9-6.5 mmol/L).

DISCUSSION
We found that compared with patients with non-DM, those with PDDM, NDDM-RH and stress hyperglycemia were related to higher risk of 90-day stroke recurrence after minor stroke or TIA in this subanalysis of the CHANCE trial. Compared with PDDM, NDDM-RH had a similar risk of 90-day stroke recurrence, while stress hyperglycemia was related to higher risk of 90-day stroke recurrence after minor AGING  However, there is no guidelines specifically definition of stress hyperglycemia and the identification of such patients is complex. Because a part of patients without a history of DM and with fasting blood glucose >7mmol/L or hypoglycemic agents were classified as NDDM-RH in our study, the prevalence of stress hyperglycemia was lower than other studies. [22]. The proportion of type 1 DM was 0.09% among patients with PDDM in our study. A population-based registry study has shown that incidence per 100000 persons years of type 1 DM was 2.68 for 10-14 years with a peak and 0.69 for ≥30 years. The incidence of type 1 DM decreased steadily with age. [23] Another cohort study enrolled Chinese adults aged from 35 to 74 years has shown that the incidence of type 2 DM was 9.6 and 9.2 per 1000 persons years for men and women, respectively. [24] Therefore, the rate of NDDM-RH caused by type 1 DM may be very low in our study, compared to those caused by type 2 DM.

AGING
DM has been verified as a distinct risk factor of recurrent stroke after ischemic stroke or TIA. [7][8][9][10]25] A previous study has revealed that NDDM-RH in Chinese patients with acute ischemic stroke was related to 1-year poor outcome. [14] PDDM and NDDM-RH were related to elevated risk of 90-day stroke recurrence in minor stroke or TIA in this study, which was in keeping with the previous research. [14] Patients with NDDM-RH have a similarly poor prognosis in acute myocardial infarction underwent an operation of percutaneous coronary intervention as those with PDDM. [26] Stress hyperglycemia has been demonstrated as a distinct risk factor for in-hospital death and 90-day stroke recurrence after ischemic stroke. [15,16] However, little research has compared the risk of 90-day stroke recurrence between PDDM, NDDM-RH, and stress hyperglycemia after minor stroke and TIA. Our study added evidence that compared with PDDM, NDDM-RH had a similar risk of 90-day stroke recurrence, while stress hyperglycemia had higher risk of 90-day stroke recurrence after minor stroke and TIA.
These results in our study were conflict with a previous study which demonstrated that in acute ischemic stroke stress hyperglycemia not distinctly associated with unfavorable outcome. [22] Acute ischemic stroke has traditionally been classified as five etiological subtypes based upon the trial of ORG 10172 in acute stroke treatment (TOAST) criteria: 1) cardioembolism, 2) large-artery atherosclerosis, 3) small-vessel occlusion, 4) stroke of other determined, and 5) stroke of AGING Several aspects may account for the phenomenon that stress hyperglycemia had higher risk of 90-day stroke AGING recurrence compared to PDDM after minor stroke or TIA. Firstly, stress hyperglycemia is relative hyperglycemia at risk of critical illness caused by neuro-hormonal derangements and inflammatory response. [32] Secondly, compared with chronic sustained hyperglycemia, fluctuations of blood glucose has a more specific striking impact on oxidative stress [33] and impairs endothelial function. [34] Acute hyperglycemia can increase circulating cytokine concentrations by an oxidative mechanism. [35] These are the critical factors that contribute to cerebral vascular events. Furthermore, a general population has shown that fluctuation of fasting blood glucose significantly increased the risk of cardiovascular diseases in the general individuals. [36] However, future large-scale cohorts are needed to explain this association.
We found that clopidogrel and aspirin was related to lower risk of 90-day stroke recurrence in non-DM and stress hyperglycemia compared with aspirin only, and these associations were not observed in patients with PDDM and NDDM-RH. However, no interaction effect of antiplatelet therapy among the four groups for the risk of 90-day stroke recurrence was observed. This may be ascribed to the small sampling size of our analysis. Caution is still required in our interpretations for lack of follow-up of dynamic glucose and glycated albumin. Future largescale studies are needed to illustrate the interaction of antiplatelet therapy by different status of glucose metabolism status.
This study had several limitations. First, since we didn't measure oral glucose tolerance tests and HbA1c, NDDM-RH was based upon fasting blood glucose, GA, and using of hypoglycemic medications during hospitalization, which may have led to misclassification of the groups. Second, patients with self-reported history of DM diagnosed by physician at admission were defined as PDDM.
Data from self-reported information was not as precise as data from detailed medical records. Nevertheless, our well-designed randomized controlled trial may make up for this deficiency to some extent. Third, the dynamic changes of GA and fasting blood glucose were not available in the CHANCE trial. Future studies with dynamic changes of those biomarkers are needed. Fourth, patients enrolled in the CHANCE was confined to acute noncardioembolic high-risk TIA or minor stroke within 24 hours. Our findings may be not inapplicable to other subtypes or moderate to severe acute ischemic stroke. Hence, the results of our study should be interpreted carefully and are needed to be confirmed in the future large-scale studies.
In conclusion, our study demonstrated that PDDM, NDDM-RH and stress hyperglycemia were related to higher risk of 90-day stroke recurrence in minor stroke or TIA. Compared with PDDM, NDDM-RH had a similar risk of 90-day stroke recurrence, while stress hyperglycemia was related to higher risk of 90-day stroke recurrence in minor stroke or TIA. Early identification and rigid management of NDDM-RH and stress hyperglycemia may help to decrease the 90-day stroke risk after minor stroke or TIA.

Study design and population
Description regarding the rationale and design of the CHANCE trial have been reported in detail. [6,37] Briefly, it was a randomized, double-blind, placebocontrolled clinical trial carried out in 114 hospitals in China from October 1, 2009 to July 30, 2012. In total, 5170 patients with non-cardioembolic minor stroke or high-risk TIA within 24 hours were randomly assigned to the treatment regimens of aspirin only or clopidogrel plus aspirin. Inclusion criteria for the CHANCE trial are summarized as follows: 1) at least 40 years of age; 2) having an acute minor ischemic stroke (NIHSS ≤ 3) or high-risk TIA (ABCD 2 ≥ 4); 3) able to take study medications within 24 hours after onset. There were 73 (64%) prespecified hospitals took part in the biomarker substudy on a voluntary basis. This biomarker substudy of CHANCE consecutively recruited 3044 patients.
Ethics approval was granted by the Ethics Committee all participating sites. All patients or their representatives provided written informed consent. CHANCE was registered with ClinicalTrials.gov (Number: NCT00979589).

Data collection
Patient baseline information including age, sex, height, weight, history of DM, hypertension, smoking status, hypercholesterolemia, ischemic stroke, TIA, atrial fibrillation or flutter, myocardial infarction, coronary heart disease, angina, and NIHSS at admission were recorded by trained and certified interviewers masked to randomization by means of face-to-face interviews. Plasma glucose measurements after overnight fasting were performed within 48 hours after admission.

Measurement of GA and hs-CRP
Fasting venous blood was collected from each fasting patient participating in the biomarker substudy within AGING 24±12 hours after randomization. Blood samples were collected by face-to-face interviews at each center and delivered through cold-chain to Beijing Tiantan Hospital and stored at −80° C. GA assay was centrally measured with a specific equipment (catalog number 4085-717; Ruiyuan Bio-Technique Co.Ltd., Ningbo, China) through a Roche Modular P800 system. We used the percentage of total serum albumin to express the levels of GA. [38,39] Hs-CRP was measured through a turbidimetric immunoassay (Ji'en Technique Co Ltd, Shanghai, China) on a Roche Modular P800 system (Roche, Basel, Switzerland). [31] All measurements were centrally conducted by laboratory technicians who were not informed of study assignments and clinical outcomes of patients.

Groups according to glucose metabolism status
Non-DM were defined if patients met all of the following criteria: (1) without a history of physiciandiagnosed DM; (2) fasting plasma glucose < 7.0 mmol/L; (3) without using hypoglycemic medications during hospitalization. PDDM was defined based on the self-reported history of physician-diagnosed DM. According to a GA level of ≥15.5% was the optimal cut point that may predict the presence of early-stage diabetes, [40] patients without a history of physiciandiagnosed DM but with fasting plasma glucose≥ 7.0 mmol/L, or used medications to decrease blood glucose levels for any reason during hospitalization were classified as NDDM-RH or stress hyperglycemia. NDDM-RH was identified if patients without a history of DM fulfilled these two inclusion criteria: (1) using medications to decrease blood sugar levels for any reason during hospitalization, or fasting plasma glucose≥ 7.0 mmol/L; [8,41] (2) GA level of ≥15.5%. Stress hyperglycemia was identified if patients without a history of DM fulfilled these two inclusion criteria: (1) using medications to decrease blood sugar levels for any reason during hospitalization, or fasting plasma glucose≥ 7.0 mmol/L; [8,41] (2) GA level of <15.5%.

Follow-up and outcome assessment
Patients were followed up by trained site coordinators at 90 days. [6] The primary efficacy outcome was a 90day stroke recurrence (including ischemic or hemorrhagic stroke). The secondary efficacy outcomes were 90-day recurrent ischemic stroke and composite events (including ischemic stroke, hemorrhagic stroke, myocardial infarction, or vascular causes of death). Safety outcome was any bleeding during 90-day follow up. Any event related to the outcomes was are adjudicated by the central adjudication committee who were not informed of the study treatment assignments.

Statistical analysis
Categorical variables were expressed as frequencies (percentage) and continuous variables were expressed as medians (interquartile ranges). Categorical variables were estimated by χ 2 test. Continuous variables were estimated with Kruskal-Wallis test. To estimate the interaction effect of glucose metabolism status by treatment assignments on the 90-day stroke recurrence, we analyzed different glucose metabolism status × treatment assignment on incident of 90-day stroke recurrence by using multivariable Cox models. We performed multivariable Cox regression models to estimate the relationship between different glucose metabolism status and outcomes. Two models were performed. In the first model, age and sex were adjusted. In the second model, all the baseline variables listed in the Table 1 were adjusted. Adjusted hazard ratios (HR) with 95 % confidence intervals (CI) were reported. Cumulative probability of 90-day stroke recurrence, recurrent ischemic stroke and composite events were constructed by Kaplan-Meier curves.
All tests were two-sided and we considered a p value less than 0.05 as statistically significance. Data analysis were done with SAS software, version 9.4 (SAS Institute Inc., Cary, NC).

AUTHOR CONTRIBUTIONS
Yuzhou Guo: Study concept and design, data analysis and interpretation, manuscript drafting; Guangyao Wang: Study concept and design, data analysis and interpretation, manuscript drafting; Jing Jing: Data acquisition, study supervision or coordination; Anxin Wang: Data analysis and interpretation, manuscript revising; Xiaoli Zhang: Data analysis and interpretation, manuscript revising; Xia Meng: Data acquisition, study supervision or coordination; Xingquan Zhao: Data acquisition, study supervision or coordination; Liping Liu: Data acquisition, study supervision or coordination; Hao Li: Study supervision, data interpretation, and commented on the drafts; David Wang: Study supervision, data interpretation, and commented on the drafts; Yongjun Wang: Study AGING concept and design, data acquisition, data analysis, obtaining funding and interpretation; Yilong Wang: Study concept and design, data analysis and interpretation obtaining funding, study supervision or coordination; Yuzhou Guo and Guangyao Wang are cofirst authors; Yilong Wang and Yongjun Wang are cocorresponding authors.

ACKNOWLEDGMENTS
We thank all participating hospitals, physicians and nurses in the CNANCE trial.

CONFLICTS OF INTEREST
The authors declare that they have no conflicts of interest.

FUNDING
The study was supported by grants from the National Natural Science Foundation of China (81825007)