Impact of glycemic gap on 30-day adverse outcomes in patients with acute ST-segment elevation myocardial infarction

doi:10.21203/rs.3.rs-1734357/v1

Download PDF

Research Article

Impact of glycemic gap on 30-day adverse outcomes in patients with acute ST-segment elevation myocardial infarction

https://doi.org/10.21203/rs.3.rs-1734357/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background: Stress-induced hyperglycemia (SIH) generally occurs in critical illness. Recently, glycemic gap (GAP) has been considered to be a superior indicator of SIH. However, data on the association between GAP and prognosis in ST-segment elevation myocardial infarction (STEMI) is limited. This observational study aimed to estimate the prognostic value of GAP_mean, defined as the difference between mean blood glucose level within 24 hours after admission and A1c-derived average glucose (ADAG), in patients with acute STEMI.

Methods: A total of 4952 patients with acute STEMI were included in the final analysis, and they were divided into four groups according to GAP_mean quartiles and diabetes mellitus (DM). The primary outcomes were defined as all-cause mortality and major adverse cardiovascular events (MACEs). Cox proportional hazards regression analysis and net reclassification improvement (NRI) analysis were performed.

Results: At 30 days of follow-up, 324 (6.5%) deaths and 569 (11.5%) MACEs occurred. With the elevation of GAP_mean, the incidence of all-cause mortality (4.0%, 5.6%, 6.5%, and 10.1%) and MACEs (7.3%, 9.6%, 11.4%, and 17.7%) significantly increased. The receiver operating characteristic curve analysis demonstrated that GAP_meanwas superior to admission blood glucose (ABG) and GAP_adm (defined as the difference between ABG and ADAG) to detect adverse outcomes. Multivariate Cox regression analysis revealed that elevated GAP_meanwas independently associated with all-cause death and MACEs, with the hazards ratio (HR) for all-cause death in the second, third, and fourth quartiles with the first quartile as reference were 1.50 (95% CI 1.04-2.18), 1.62 (95% CI 1.13-2.32), and 2.18 (95% CI 1.55-3.08), and for MACEs were 1.37 (95% CI 1.04-1.81), 1.56 (95% CI 1.20-2.04), and 2.13 (95% CI 1.65-2.76), respectively, which was independent of the status of DM. Continuous NRI analysis revealed that GAP_mean significantly improved risk stratification for all-cause mortality and CEEs by 21.6% and 19.8%, respectively.

Conclusion: Glycemic gap between the mean level of blood glucose within 24 hours after admission and ADAG was independently associated with 30-day all-cause death and MACEs in patients with acute STEMI, which was consistent in both DM patients and non-DM patients, and it can provide incremental accuracy in the risk stratification of STEMI.

ST-segment elevation myocardial infarction

Glycemic gap

Stress-induced hyperglycemia

All-cause mortality

Major adverse cardiovascular events

Ischemic heart disease (IHD) is a leading cause of death and disability worldwide, which affects approximately 126 million individuals and leads to 9 million deaths globally according to data of the Global Burden of Disease study, causing enormous medical costs and public health burden [1, 2].ST-segment elevation myocardial infarction (STEMI) is the most acute and critical manifestation of IHD, and leads to substantial morbidity and mortality. Early reperfusion therapy of thrombolysis or percutaneous coronary intervention (PCI) is proven to notably improve the prognosis of STEMI by preserving the viability of the ischemic myocardium and limiting infarct size, and recommended by current guidelines [3].

Risk stratification is of great importance for the prognostic evaluation of patients with STEMI. TIMI (Thrombolysis In Myocardial Infarction) risk score (TRS) composed of 7 items derived from clinical characteristics of patients with acute coronary syndrome has been validated in the risk stratification of STEMI, and widely applied in clinical practice [4]. In critical illness, the production of counter-regulatory hormones including cortisol, catecholamines, and glucagon notably increases, which mediates accelerated gluconeogenesis and leads to the occurrence of stress-induced hyperglycemia (SIH) [5]. SIH has been recognized as an important indicator of the severity of diseases, as it is closely associated with poor prognosis in a wide variety of pathologies [6-8]. Compared with admission blood glucose (ABG), glycemic gap, derived by subtracting A1c-derived average glucose (ADAG) from blood glucose levels, has been identified to be a superior indicator of SIH as it improves the accuracy of assessment by removing the impact of chronic hyperglycemia on the evaluation of disease severity [9-11]. However, the predictive value of glycemic gap in STEMI has not been fully investigated, and it is speculated that it can provide incremental accuracy in the risk stratification of acute STEMI on the basis of TRS. This present study aimed to evaluate the association between GAP_mean, defined as the difference between mean blood glucose level (MGL) within 24 hours after admission and ADAG, and 30-day adverse outcomes in patients admitted with acute STEMI, and to assess the incremental predictive value of GAP_mean in the risk stratification of STEMI.

Study Design and Participants

The study was designed as a multi-center observational study and conducted across 274 centers in China. A total of 7469 patients with a diagnosis of acute STEMI admitted within 12 hours from symptom onset between June 2001 to July 2004 were recruited consecutively. The diagnosis of acute STEMI was based on the following criteria: (1) typical chest pain or equivalent symptoms; (2) dynamic changes of electrocardiogram (ECG) with ST-segment elevation of more than two adjacent leads (≥ 0.1 mV in extremity leads and ≥ 0.2 mV in precordial leads) or new left bundle branch block (LBBB); (3) elevation of biochemical markers of myocardial injury, including creatine kinase-MB and troponin I. Patients with contraindications of anticoagulant therapy, hemorrhagic stroke within 12 months, malignancies and life-threatening diseases with expected survival less than 1 month, pregnancy, and inability to complete expected follow-up were excluded from the study. All participants provided written informed consent. The study protocol was approved by the Ethics Committee of each center and complied with the declaration of Helsinki.

After admission, all patients received treatments as far as possible to in accordance with guideline recommendations for the management of STEMI. Revascularization therapy of thrombolysis or percutaneous coronary intervention (PCI) was performed according to the clinical circumstance and healthcare level of each center. In the present study, 4952 patients were included in the final analysis after excluding 10 patients without detailed data, 2042 patients lacking laboratory results of HbA1c, and 465 patients lacking complete blood glucose values for three times within 24 hours (Figure 1).

Data Collection and Laboratory Measurements

Baseline information of participants including age, sex, weight, blood pressure (BP), heart rate, medical histories including myocardial infarction (MI), heart failure (HF), stroke, hypertension, and diabetes mellitus (DM), revascularization therapy including thrombolysis and PCI, and medications including antiplatelets, β-blockers, statins, angiotensin-converting enzyme inhibitors (ACEIs) or angiotensin receptor antagonists (ARBs), calcium channel blockers (CCBs), nitrates, and diuretics were extracted from the medical records. Admission vital signs were accurately measured in supine position after at least 3 minutes of rest. Location of MI on ECG and Killip classification were also recorded. The admission TRS was calculated by assigning 1 point each for the existence of weight < 67 kg, anterior ST-segment elevation (STE) or LBBB, and time to treatment > 4 hours, 2 points each for the existence of age ≥ 65 years and < 75 years, heart rate > 100 bpm, and Killip classification II-IV, and three points for the existence of age ≥ 75 years and systolic blood pressure (SBP) < 100 mmHg [4]. Laboratory data including triple venous blood glucose values within 24 hours and HbA1c values on admission were carefully collected. All medical records of patients were reviewed with any dispute in the process of data collection resolved by discussing with senior investigators.

For laboratory tests, venous blood samples were routinely drawn from an antecubital vein of participants using a 21-gauge sterile syringe into EDTA-treated tubes or plain tubes. The venous blood glucose level was tested on admission, at 6 ± 2 hours and 24 ± 2 hours after admission for each participant using the glucose oxidase (GOD) method. For the test of HbA1c, blood samples were collected on admission from the cubital vein and transported to the central laboratory of Fuwai Hospital located in Beijing, and the high-performance liquid chromatography method was utilized. All centers followed the same standard blood collection procedures and testing methods.

ADAG, which represented chronic average blood glucose levels within 3 months, was calculated for each participant according to the following formula: ADAG (mmol) = (HbA1c * 1.59) – 2.59 [12]. MGL was calculated as the mean of three blood glucose values within 24 hours after admission of each participant. Standard deviation (SD) of MGL was calculated using the following formula: SD = squr [((ABG – MGL)² + (blood glucose at 6 ± 2 hours – MGL)² + (blood glucose at 24 ± 2 hours – MGL)²) / 2], and coefficient of variation (CV) was calculated as MGL divided by SD for each participant, which reflected the fluctuation of blood glucose within 24 hours. GAP_adm, defined as the difference between ABG and ADAG, was calculated as follows: GAP_adm (mmol) = ABG – ADAG, and GAP_mean was calculated as follows: GAP_mean (mmol) = MGL – ADAG. DM was defined as HbA1c levels ≥ 6.5%, and non-DM was defined as HbA1c levels < 6.5%.

Follow-Up and Outcomes

All participants were follow-up for 30 days by trained research personnel through interviewing patients in clinic, telephoning patients or their relatives, or checking their medical records. The primary outcomes were defined as all-cause mortality and major adverse cardiovascular events (MACEs) at 30 days after enrollment. MACEs were composite outcomes of all-cause mortality, reinfarction, cardiogenic shock, cardiac arrest, stroke, and major bleeding. All-cause mortality included cardiovascular death (defined as any death with a cardiovascular cause) and non-cardiovascular death (defined as deaths due to a clearly documented non-cardiovascular cause). Reinfarction was defined as recurrent typical chest pain with new ischemic electrocardiographic changes (ST re-elevation or depression or new Q waves) and further increased enzyme levels (to twice the upper limit of normal if it had returned to baseline or if already elevated, with a further elevation by 50%). Cardiogenic shock was defined as persistent hypotension (SBP < 90 mmHg) that did not respond to fluid titration and required intravenous inotropic therapy or intra-aortic balloon pump. Cardiac arrest was defined as successful resuscitation of ventricular fibrillation or persistent ventricular tachycardia. Stroke was defined as an acute focal neurological deficit in a location consistent with the territory of a major cerebral artery lasting > 24 hours or is interrupted by death within 24 hours as verified by computed tomography or magnetic resonance imaging. Major bleeding included life-threatening bleeding, and/or symptomatic bleeding in a vital area or organ such as intracranial bleeding, and/or bleeding leading to surgical intervention, and/or overt bleeding causing a fall in hemoglobin level of ≥ 20 g/L (1.24 mmol/L) or leading to transfusion of ≥ 2 units of whole blood or packed red cells. The outcomes were evaluated by well-trained research personnel who were blinded to the study objective.

Statistical Analysis

All participants were divided into 4 groups according to the quartiles of GAP_mean (Q1: < -0.45 mmol/L, Q2: -0.45-0.90 mmol/L, Q3: 0.90-2.72 mmol/L, and Q4: ≥ 2.72 mmol/L). Continuous variables were expressed as means ± SDs for normally distributed data or quartiles for non-normally distributed data, with the normality of the distribution evaluated using the Kolmogorov-Smirnov test. Categorical variables were expressed as frequencies (percentage). Comparisons between groups were conducted using analysis of variance or the Mann-Whitney U test for continuous variables and the Pearson’s chi-square test or Fisher’s exact test for categorical variables. The receiver operating characteristic curve (ROC) method was performed to compare the accuracy of different glycemic indicators including ABG, GAP_adm, and GAP_mean in the detection of outcomes. The Kaplan-Meier method was utilized to construct cumulative survival and MACE-free curves, with between-group comparisons conducted using the Log-rank test. To evaluate the independent association between GAP_meanand outcomes, multivariate Cox proportional hazards regression analyses were conducted by adjusting for age, sex, weight, heart rate, SBP, anterior STE or LBBB, Killip classification II-IV, time to treatment > 4 hours, MI, stroke, HF, DM, hypertension, PCI, thrombolysis, antiplatelets, statins, β-blockers, and ACEI/ARBs. The enter method was used to obtain the hazard ratio (HR) and 95% confidence interval (CI) of all variables in the models, with entry criterion set up at P < 0.05 and exit criterion set up at P = 0.10. Net reclassification improvement (NRI) and integrated discrimination improvement (IDI) analyses were performed to quantify the incremental value of GAP_mean in the risk stratification of STEMI. In general, using the survival data of this study, increased discriminative values of GAP_mean were calculated by adding GAP_mean to the models constructed using all other variables included in multivariate Cox regression models, with the expected numbers of event and nonevent subjects computed using Kaplan-Meier survival estimates at a follow-up duration of 30 days. Confidence intervals and P values were estimated using 300 bootstrap replications [13, 14].

All statistical analysis was conducted using SPSS software package (version 23.0; IBM) and R software (version 3.3.3; R Foundation for Statistical Computing), and P < 0.05 with 2-tailed was considered significant.

Baseline characteristics

Baseline characteristics of participants according to GAP_mean quartiles are shown in Table 1. The average age of overall patients was 62.6 years, and 29.5% of patients were female. Compared with patients with low GAP_mean values, patients with high GAP_mean values were older, thinner, had faster heart rate, anterior STE or LBBB, worse Killip classification and higher TRS, besides, they were more likely to be female, and suffer from hypertension, DM, and stroke. Values of BG, MGL, SD, CV, and GAP_adm were significantly greater in patients with higher GAP_mean values. With regards to treatment, patients in high GAP_mean quartiles received reperfusion therapy, antiplatelets, diuretics, and insulin more frequently, while they were less likely to use nitrates compared with patients in low GAP_mean quartiles.

Adverse outcomes according to GAP_mean quartiles and the status of DM

Adverse outcomes according to GAP_mean and the status of DM are shown in Figure 2. Over the follow-up of 30 days, a total of 324 (6.5%) deaths and 569 (11.5%) MACEs occurred. The incidence of all-cause death and MACEs were significantly increased with the elevation of GAP_mean quartiles, with the mortality of 4.0%, 5.6%, 6.5%, and 10.1%, and the incidence rate of MACEs of 7.3%, 9.6%, 11.4%, and 17.7% in each GAP_mean quartile, respectively. This trend was roughly observed in non-DM group and DM group, besides, patients in DM group had worse prognosis compared with those in non-DM group. ROC analysis indicated that GAP_meanwas the best glycemic indicator to detect adverse outcomes, with the area under the curve of 0.614 and 0.611 for all-cause mortality and MACEs, respectively (Supplementary File 1). The Kaplan-Meier survival curves according to GAP_mean quartiles are shown in Figure 3. All-cause mortality and the incidence of MACEs were significantly higher in patients in high GAP_mean quartiles.

Associations between GAP_mean and adverse outcomes

In multivariate Cox regression models in which GAP_mean served as continuous variable, per 1 mmol/L increment in GAP_mean was independently associated with 10% increased risk of all-cause death (HR 1.10, 95% CI 1.07-1.14, P < 0.001) and 9% increased risk of MACEs (HR 1.09, 95% CI 1.06-1.11, P < 0.001). Besides, older age, lower SBP, faster heart rate, Killip classification II-IV, time to treatment > 4 hours, prior stroke were associated with increased risk of all-cause death and MACEs, while treatment of PCI, antiplatelets, statins, β-blockers, ACEI/ARBs were related to reduced risk of all-cause death and MACEs. Female and anterior STE or LBBB were identified to be risk factors of all-cause mortality, while low body weight and receiving thrombolysis were associated with increased risk of MACEs (Table 2).

Figure 4 shows the associations of GAP_mean quartiles with all-cause death and MACEs according to the status of DM. After adjusting for confounding factors in multivariate Cox regression models, increased GAP_meanquartiles were independently associated with all-cause death and MACEs. The HRs for all-cause death in the second, third, and fourth quartiles compared with the first quartile were 1.50 (95% CI 1.04-2.18), 1.62 (95% CI 1.13-2.32), and 2.18 (95% CI 1.55-3.08), respectively, and for MACEs were 1.37 (95% CI 1.04-1.81), 1.56 (95% CI 1.20-2.04), and 2.13 (95% CI 1.65-2.76), respectively. Moreover, the status of DM did not make significant difference to the association between GAP_mean and adverse outcomes.

Incremental value of GAP_mean in the risk stratification of STEMI

Table 3 shows results of NRI and IDI analyses assessing the adding value of GAP_meanin the risk stratification of STEMI. Variables including age, sex, weight, SBP, heart rate, anterior STE or LBBB, Killip classification II-IV, time to treatment > 4 hours, hypertension, MI, DM, stroke, HF, thrombolysis, PCI, antiplatelets, statins, β-blockers, and ACEI/ARBs were utilized to construct the original models for predicting all-cause mortality and MACEs, and its discrimination and calibration were confirmed to be good (supplementary file 2). NRI and IDI analyses revealed that the addition of GAP_mean to the original models provided significant improvements in risk stratification for both all-cause mortality and MACEs. The continuous NRI was 0.216 (95% CI 0.104-0.328) for all-cause mortality and 0.198 (95% CI 0.111-0.286) for MACEs, and IDI was 0.013 (95% CI 0.005-0.020) for all-cause death and 0.010 (95% CI 0.005-0.015) for MACEs.

In the present study, we assessed the predictive value of GAP_mean on adverse outcomes in patients with acute STEMI, and the major findings of our analysis are as follows. First, patients with different GAP_mean levels exhibited distinct clinical characteristics, and patients with high GAP_mean values were older, presented with more severe clinical manifestations, had more comorbidities and greater blood glucose fluctuations than patients with low GAP_mean values. Second, a significantly worse prognosis was noted in patients in high GAP_mean quartiles compared with those in low GAP_mean quartiles, and this trend was roughly observed in both DM group and non-DM group. ROC analysis indicated that GAP_mean was superior than other glycemic parameters including ABG and GAP_adm. Third, multivariate Cox regression analysis revealed that GAP_mean was independently associated with increased risk of all-cause death and MACEs, which was not affected by the status of DM. Finally, the results of NRI and IDI analyses further revealed that the addition of GAP_mean to original models constructed by other predictors significantly improved the predictive efficiency of models, confirming the incremental value of GAP_mean in the risk stratification of acute STEMI.

Stress hyperglycemia generally refers to a transient increase in blood glucose levels as the result of sympathetic system activation during critical state, and has been identified as an important indicator of the severity of multiple diseases and states including sepsis, surgery, trauma, acute illness requiring intensive critical care, acute ischemic stroke, intracerebral hemorrhage, etc. [4–7]. Glucose concentrations at admission has long been used as the indicator of stress hyperglycemia, and previous studies have revealed its close relationship with adverse outcomes. Notably, this association seems to be more significant in non-diabetic patient, by comparison, the impact of elevated blood glucose at admission on outcomes was markedly weakened in patients with DM [15–17]. ABG may not authentically reflect the stress-induced increase in blood glucose levels, as it can be interfered by the status of DM. Thus, ABG may not be an ideal indicator of SIH as it is unable to distinguish whether the increase in blood glucose is caused by DM or acute physiological stress. Recently, glycemic gap, defined as the difference between blood glucose after admission and ADAG, is identified as a novel glycemic indicator and can reflect stress hyperglycemia more accurately by eliminating the influence of chronic hyperglycemia compared with ABG. The prognostic value of glycemic gap has been investigated in a variety of diseases, and it has exhibited strong correlations with disease severity and prognosis [8–10, 18–20]. Dorn et al. explored admission factors and outcomes in relation to GAP_adm in patients with intracerebral hemorrhage (ICH), and found that elevated GAP_adm was associated with worse Glasgow Goma Scale and ICH scores on admission, as well as in hospital mortality and poor discharge status [8], which was also confirmed in diabetic patients with ICH [18]. Kim et al. assessed the effect of admission-glucose gap on short-term functional outcomes in 1332 patients with acute ischemic stroke, and found that elevated admission-glucose gap was related to high modified Rankin Scale score at 3 months after stroke onset [19]. Jensen et al. evaluated the impact of glycemic gap on prognosis in 1933 patients with community-acquired pneumonia, and concluded that patients with the highest and lowest glycemic gap had increased risk of 90-day mortality, and the association was not modified by DM [10]. In critically ill patients with DM, glycemic gap between MGL within the first 7 days after admission to ICU and ADAG was identified to be independently related with 28-day mortality [9], consistent with the findings of a recent Korean study [20].

A few studies have focused on the prognostic value of glycemic gap in cardiovascular diseases. Liao et al. evaluated the impact of glycemic gap on outcomes in 425 diabetic patients hospitalized for acute HF, and found that glycemic gap > 43 mg/dl was related to increased risk of in-hospital mortality [21]. A recent study revealed that augmented glycemic gap was a marker for an increased risk of post-infarct left ventricular systolic dysfunction, which was consistent in both DM patients and non-DM patients [22]. In the field of MI, a study explored the association between glycemic gap and outcomes in 331 diabetic patients presenting with acute MI, and found that glycemic gap was superior in discriminative power regarding the development of MACEs compared with ABG, and can be used as an adjunct parameter to assess the severity and prognosis of acute MI [23]. However, the sample size of this study was small and only acute MI patients complicated with diabetes were included. To the best of our knowledge, our present study represents the first comprehensive analysis evaluating the impact of GAP_mean on 30 days outcomes in both DM and non-DM patients with STEMI. Our analysis of 4952 patients with STEMI reveals that elevated GAP_mean was significantly associated with increased risk of all-cause death and MACEs during 30-day follow-up. Our patients in high GAP_mean quartiles suffered from more comorbidities and had greater blood glucose fluctuations than patients in low GAP_mean quartiles, which was consistent with previous findings in patients with critical ill, suggesting that patients with elevated GAP_mean values were in worse status and more vulnerable [9]. Moreover, we also compared the predictive ability of multiple glycemic parameters, and found that GAP_mean was superior to ABG and GAP_adm, with the largest area under the curve in ROC analysis. This can be explained by the fact that GAP_mean eliminates the effect of basal glucose levels on the assessment of disease severity, and that MGL after admission provides more comprehensive data on acute glycemia, and thus GAP_mean can reflect the elevated blood glucose caused by stress more accurately. Our multivariate Cox regression analysis showed that GAP_mean was independently associated with poor prognosis, and consistent results were observed in both non-DM patients and DM patients when patients were grouped according to the status of DM, which also verifies that GAP_mean does eliminate the impact of basal blood glucose level on prognosis prediction, and is a powerful indicator that can reflect SIH accurately. More importantly, further NRI and IDI analyses suggests that the inclusion of GAP_mean in models constructed by other identified predictors significantly improved the predictive ability of original models for both all-cause death and MACEs, providing incremental value in the risk stratification of STEMI.

Previous studies have demonstrated that a variety of complex mechanisms are involved in the development of SIH [24, 25]. The underlying illness may result in an excessive release of hormones such as catecholamines, growth hormone, and cortisol, and cytokines such as tumor necrosis factor-α and interleukin 1, which form an intricate regulatory network of feedforward and feedback mechanisms, and this neurohormonal environment leads to abundant hepatic glucose production through gluconeogenesis and insulin resistance by blocking post-receptor insulin signaling and downregulating glucose transporter-4 and reducing insulin-medicated glucose uptake and glycogen synthesis. Hyperglycemia in turn promotes the production of cytokines, inflammatory, and oxidative stress, thus forming a vicious circle whereby hyperglycemia leads to further hyperglycemia [24]. In addition, insulin resistance can exacerbate lipolysis, and cause abundant circulating free fatty acids (FFA), aggravating inflammatory, insulin resistance and gluconeogenesis [25]. The mechanism of adverse outcomes caused by stress hyperglycemia in acute STEMI remains to be elucidated, and endothelial dysfunction, oxidative stress, altered FFA metabolism, and procoagulant effect mediated by hyperglycemia have been considered to be significant mechanisms [26]. Cardiac glucotoxicity can impact myocardial function by altering both cardiomyocyte and endothelial cell function, which are mediated by the activation of the polyol pathway and cytosolic protein kinase C, and the generation of advanced glycation end products.

A few trials have investigated the optimal glycemic target for patients with acute MI, and obtained inconsistent results [27–30]. In the DIGAMI trial, 620 AMI patients with DM or a blood glucose level > 11 mmol/L without DM were randomized into insulin arm and control arm, and all-cause mortality were significantly reduced in insulin arm at 1- and 3.4-year follow up points [27]. However, the following DIGAMI 2 trial failed to show a significant reduction in mortality in insulin-based intensive glucose control [28, 29]. A recent randomized trial BIOMArCS-2 found that intensive glucose control in the early phase of acute MI resulted in persistent harmful effects [30]. Discrepancies in these results can be explained by differences in the definition of SIH and target glucose, as well as diversity of recruited participants. Our present study revealed that GAP_mean is a superior indicator of SIH than ABG as it avoids the impact of the status of DM. It indicates that GAP_mean should be used as the parameter to reflect stress hyperglycemia in the future research exploring the efficacy and safety of intensive glucose control and establishing the optimal target blood glucose in patients with acute MI. Our findings enriched the risk stratification of adverse outcomes in STEMI patients both with DM and non-DM, which had certain guiding significance for the management of these patients in clinical practice, and it might be rational that more intensive management should be applied to these high-risk patients so as to improve their survival.

The most protruding strength of the present study is that it is a large-scale multi-center observational study with 4952 participants from 274 centers all over China to explore the association between GAP_mean and adverse outcomes in patients with acute STEMI, which to our knowledge, is the first study to comprehensively evaluate the prognostic value of GAP_mean in STEMI, with subgroup analysis according to the status of DM. In addition, data recorded in our study was completed by professionally trained clinician and researchers, with disputes during data collection resolved by discussion with senior investigators, which may increase the credibility of our data and the reliability of our results. However, several limitations should be noted in our study. First, our findings cannot demonstrate a definite causal association between elevated GAP_mean and adverse outcomes due to inherent limitations of retrospective study. Second, only a few patients received revascularization therapy in our study, thus, our conclusions should be treated with caution when extrapolating to patients with STEMI in current medical setting. Third, although we recorded the data on whether insulin was utilized or not, information on insulin usage, dosage, and modality of administration were lacking. Last, factors that might impact on blood glucose including diet and medications such as vasoactive drugs and antibiotics were not recorded in this study, and their confounding effects on the association cannot be eliminated. Therefore, large-scale and well-designed studies are required to confirm our findings.

The present study revealed that an elevated glycemic gap between the MGL within 24 hours after admission and ADAG was independently associated with 30-day all-cause mortality and MACEs in patients with STEMI, independent of the status of DM. The predictive value of GAP_mean was superior to ABG and GAP_adm。Moreover, GAP_mean can provide incremental prognostic information beyond that provided by conventional risk factors, which can be considered to be included in the risk stratification of STEMI.

ABG: admission blood glucose; ACEIs: angiotensin-converting enzyme inhibitors; ADAG: A1c-derived average glucose; ARBs: angiotensin receptor antagonists; BP: blood pressure; CCBs: calcium channel blockers; CI: confidence interval; CV: coefficient of variation; DM: diabetes mellitus; ECG: electrocardiogram; FFA: free fatty acids; GOD: glucose oxidase; HF: heart failure; HR: hazard ratio; ICH: intracerebral hemorrhage; IDI: integrated discrimination improvement; IHD: ischemic heart disease; LBBB: left bundle branch block; MACEs: major adverse cardiovascular events; MGL: mean blood glucose level; MI: myocardial infarction; NRI: net reclassification improvement; PCI: percutaneous coronary intervention; ROC: receiver operating characteristic curve; SBP: systolic blood pressure; SD: standard deviation; SIH: stress-induced hyperglycemia; STE: ST-segment elevation; STEMI: ST-segment elevation myocardial infarction; TIMI: Thrombolysis In Myocardial Infarction; TRS: Thrombolysis In Myocardial Infarction risk score.

Ethics approval and consent to participate

The study protocol was complied with the declaration of Helsinki and approved by the Ethics Committee of each center. All participants provided written informed consent.

Consent for publication

The manuscript was approved by all authors for publication.

Availability of data and materials

The datasets analyzed in the current study are available from the corresponding author upon reasonable request.

Competing interests
All authors declared no potential competing interests with respect to the research, authorship, and/or publication of this manuscript.

Funding

The present study was funded by the Capital’s Funds for Research and Application of Clinical Diagnosis and Treatment Technology (No. Z191100006619121).

Authors’ contributions

YMY and JZ contributed to study design and quality control. SW, WX, LLW, SQL, JW, XHS, and HZ contributed to data collection and clinical follow-up. SW performed statistical analysis and wrote the initial draft of the manuscript. YMY and JZ reviewed and revised the manuscript, and made the final approval of the version to be published. All authors reviewed and approved the final manuscript.

Acknowledgements

We thank all the participants in the study and the members of the survey teams for their contributions to this study, as well as the groups providing financial support.

Khan MA, Hashim MJ, Mustafa H, Baniyas MY, Al Suwaidi SKBM, AlKatheeri R, et al. Global Epidemiology of Ischemic Heart Disease: Results from the Global Burden of Disease Study. Cureus. 2020;12(7):e9349.
Virani SS, Alonso A, Benjamin EJ, Bittencourt MS, Callaway CW, Carson AP. Heart Disease and Stroke Statistics-2020 Update: A Report From the American Heart Association. Circulation. 2020.141(9):e139-596.
Ibanez B, James S, Agewall S, Antunes MJ, Bucciarelli-Ducci C, Bueno H, et al. 2017 ESC Guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation: The Task Force for the management of acute myocardial infarction in patients presenting with ST-segment elevation of the European Society of Cardiology (ESC). Eur Heart J. 2018;39(2):119–77.
Antman EM, Cohen M, Bernink PJ, McCabe CH, Horacek T, Papuchis G, et al. The TIMI risk score for unstable angina/non-ST elevation MI: A method for prognostication and therapeutic decision making. JAMA. 2000;284(7):835–42.
Mifsud S, Schembri EL, Gruppetta M. Stress-induced hyperglycaemia. Br J Hosp Med (Lond). 2018;79(11):634–9.
Vanhorebeek I, Gunst J, Van den Berghe G. Critical Care Management of Stress-Induced Hyperglycemia. Curr Diab Reports. 2018;18(4):17.
Li M, Chen G, Feng Y, He X. Stress Induced Hyperglycemia in the Context of Acute Coronary Syndrome: Definitions, Interventions, and Underlying Mechanisms. Front Cardiovasc Med. 2021;8:676892.
Li WA, Moore-Langston S, Chakraborty T, Rafols JA, Conti AC, Ding Y. Hyperglycemia in stroke and possible treatments. Neurol Res. 2013;35(5):479–91.
Dorn AY, Sun PY, Sanossian N, Nguyen PL, Emanuel BA, Kim-Tenser MA, et al. Admission glycemic gap in the assessment of patients with intracerebral hemorrhage. Clin Neurol Neurosurg. 2021;208:106871.
Lou R, Jiang L, Zhu B. Effect of glycemic gap upon mortality in critically ill patients with diabetes. J Diabetes Investig. 2021;12(12):2212–20.
Jensen AV, Baunbæk Egelund G, Bang Andersen S, Petersen PT, Benfield T, Witzenrath M, et al. The Glycemic Gap and 90-Day Mortality in Community-acquired Pneumonia. A Prospective Cohort Study. Ann Am Thorac Soc. 2019;16(12):1518–26.
Lenters-Westra E, Slingerland RJ. Hemoglobin A1c determination in the A1C-Derived Average Glucose (ADAG) study. Clin Chem Lab Med. 2008;46(11):1617–23.
Pencina MJ, D'Agostino RB Sr, Steyerberg EW. Extensions of net reclassification improvement calculations to measure usefulness of new biomarkers. Stat Med. 2011;30(1):11–21.
Uno H, Tian L, Cai T, Kohane IS, Wei LJ. A unified inference procedure for a class of measures to assess improvement in risk prediction systems with survival data. Stat Med. 2013;32(14):2430–42.
Kosiborod M, Rathore SS, Inzucchi SE, Masoudi FA, Wang Y, Havranek EP, et al. Admission glucose and mortality in elderly patients hospitalized with acute myocardial infarction: implications for patients with and without recognized diabetes. Circulation. 2005;111(23):3078–86.
Cheung NW, Li S, Ma G, Crampton R. The relationship between admission blood glucose levels and hospital mortality. Diabetologia. 2008;51(6):952–5.
Itzhaki Ben Zadok O, Kornowski R, Goldenberg I, Klempfner R, Toledano Y, Biton Y, et al. Admission blood glucose and 10-year mortality among patients with or without pre-existing diabetes mellitus hospitalized with heart failure. Cardiovasc Diabetol. 2017;16(1):102.
Zarean E, Lattanzi S, Looha MA, Napoli MD, Chou SH, Jafarli A, et al. Glycemic Gap Predicts in-Hospital Mortality in Diabetic Patients with Intracerebral Hemorrhage. J Stroke Cerebrovasc Dis. 2021;30(5):105669.
Kim Y, Lee SH, Kim C, Kang MK, Yoon BW, Kim TJ, et al. Personalized Consideration of Admission-Glucose Gap between Estimated Average and Initial Glucose Levels on Short-Term Stroke Outcome. J Pers Med. 2021;11(2):139.
Ha EY, Chung SM, Park IR, Lee YY, Choi EY, Moon JS. Novel Glycemic Index Based on Continuous Glucose Monitoring to Predict Poor Clinical Outcomes in Critically Ill Patients: A Pilot Study. Front Endocrinol. 2022;13:869451.
Liao WI, Lin CS, Lee CH, Wu YC, Chang WC, Hsu CW, et al. An Elevated Glycemic Gap is Associated with Adverse Outcomes in Diabetic Patients with Acute Myocardial Infarction. Sci Rep. 2016;6:27770.
Zhu Y, Liu K, Meng S, Jia R, Lei X, Chen M, et al. Augmented glycaemic gap is a marker for an increased risk of post-infarct left ventricular systolic dysfunction. Cardiovasc Diabetol. 2020;19(1):101.
Wang CH, Chang JL, Huang CH, Chang WT, Tsai MS, Yu PH, et al. The association between long-term glycaemic control, glycaemic gap and neurological outcome of in-hospital cardiac arrest in diabetics: A retrospective cohort study. Resuscitation. 2018;133:18–24.
Dungan KM, Braithwaite SS, Preiser JC. Stress hyperglycaemia. Lancet. 2009;373(9677):1798–807.
Papachristoforou E, Lambadiari V, Maratou E, Makrilakis K, et al. Association of Glycemic Indices (Hyperglycemia, Glucose Variability, and Hypoglycemia) with Oxidative Stress and Diabetic Complications. J Diabetes Res. 2020:7489795.
Scheen M, Giraud R, Bendjelid K. Stress hyperglycemia, cardiac glucotoxicity, and critically ill patient outcomes current clinical and pathophysiological evidence. Physiol Rep. 2021;9(2):e14713.
Malmberg K, Rydén L, Efendic S, Herlitz J, Nicol P, Waldenström A, et al. Randomized trial of insulin-glucose infusion followed by subcutaneous insulin treatment in diabetic patients with acute myocardial infarction (DIGAMI study): effects on mortality at 1 year. J Am Coll Cardiol. 1995;26(1):57–65.
Malmberg K, Rydén L, Wedel H, Birkeland K, Bootsma A, Dickstein K, et al. Intense metabolic control by means of insulin in patients with diabetes mellitus and acute myocardial infarction (DIGAMI 2): effects on mortality and morbidity. Eur Heart J. 2005;26(7):650–61.
Mellbin LG, Malmberg K, Norhammar A, Wedel H, Rydén L; DIGAMI 2 Investigators. Prognostic implications of glucose-lowering treatment in patients with acute myocardial infarction and diabetes: experiences from an extended follow-up of the Diabetes Mellitus Insulin-Glucose Infusion in Acute Myocardial Infarction (DIGAMI) 2 Study. Diabetologia. 2011;54(6):1308–17.
de Mulder M, Umans VA, Cornel JH, van der Zant FM, Stam F, Oemrawsingh RM, et al. Intensive glucose regulation in hyperglycemic acute coronary syndrome: results of the randomized BIOMarker study to identify the acute risk of a coronary syndrome-2 (BIOMArCS-2) glucose trial. JAMA Intern Med. 2013;173(20):1896–904.

Table 1 Baseline characteristics of participants

Variables	All patients (n = 4952)	Quartiles of GAP_mean (mmol/L)				P value
Variables	All patients (n = 4952)	Q1: < -0.45	Q2: -0.45-0.90	Q3: 0.90-2.72	Q4: ≥ 2.72
Clinical features
Age, years	62.6 ± 11.8	62.1 ± 12.1	61.7 ± 11.9	62.6 ± 11.9	63.9 ± 11.1	< 0.001
Female, n (%)	1462 (29.5)	327 (26.4)	304 (24.5)	377 (30.5)	454 (36.6)	< 0.001
Weight, kg	66.6 ± 11.9	67.7 ± 12.2	66.6 ± 12.2	66.9 ± 11.7	65.4 ± 11.3	< 0.001
Systolic blood pressure, mmHg	126.3 ± 25.5	127.4 ± 23.3	126.0 ± 25.1	126.3 ± 24.7	125.6 ± 28.5	0.326
Diastolic blood pressure mmHg	78.9 ± 16.2	79.1 ± 14.8	78.7 ± 15.6	79.3 ± 15.8	78.5 ± 18.4	0.587
Heart rate, beats/min	76.9 ± 18.1	76.1 ± 17.3	76.2 ± 16.8	76.7 ± 17.4	78.7 ± 20.8	0.001
Killip classification 1 2 3 4	4110(83.0) 668 (13.5) 93 (1.9) 81 (1.6)	1053 (85.1) 155 (12.5) 17 (1.4) 13 (1.1)	1075 (86.8) 136 (11.0) 19 (1.5) 9 (0.7)	1022 (82.7) 175 (14.2) 20 (1.6) 19 (1.5)	960 (77.5) 202 (16.3) 37 (3.0) 40 (3.2)	< 0.001
Anterior STE or LBBB	2549 (51.5)	655 (52.9)	654 (52.8)	643 (52.0)	597 (48.2)	0.060
Hypertension, n (%)	1991 (40.2)	434 (35.1)	452 (36.5)	526 (42.6)	579 (46.7)	< 0.001
Diabetes mellitus, n (%)	563 (11.4)	53 (4.3)	61 (4.9)	135 (10.9)	314 (25.3)	< 0.001
Myocardial infarction, n (%)	399 (8/1)	95 (7.7)	91 (7.3)	118 (9.5)	95 (7.7)	0.168
Heart failure, n (%)	119 (2.4)	19 (1.5)	30 (2.4)	32 (2.6)	38 (3.1)	0.090
Stroke, n (%)	467 (9.4)	79 (6.4)	106 (8.6)	109 (8.8)	173 (14.0)	< 0.001
TIMI class 1 2 3	2211 (44.6) 1972 (39.8) 769 (15.5)	598 (48.3) 475 (38.4) 165 (13.3)	597 (48.2) 487 (39.3) 155 (12.5)	537 (43.4) 507 (41.0) 192 (15.5)	479 (38.7) 503 (40.6) 257 (20.7)	< 0.001
Laboratory tests
BG on admission (mmol/L)	8.6 ± 4.2	6.2 ± 1.8	7.0 ± 1.8	8.4 ± 2.8	12.6 ± 5.7	< 0.001
BG at 6 ± 2 hours (mmol/L)	8.8 ± 4.3	6.0 ± 1.7	6.9 ± 1.9	8.6 ± 2.6	13.6 ± 5.2	< 0.001
BG at 24 ± 2 hours (mmol/L)	7.8 ± 3.5	5.8 ± 1.5	6.4 ± 1.6	7.7 ± 2.3	11.2 ± 4.7	< 0.001
MGL (mmol/L)	8.4 ± 3.3	6.0 ± 1.3	6.8 ± 1.2	8.3 ± 1.7	12.5 ± 3.5	< 0.001
SD (mmol/L)	1.5 (0.7-2.6)	0.8 (0.4-1.4)	1.1 (0.6-1.7)	1.7 (1.0-2.6)	3.3 (2.1-5.3)	< 0.001
CV (%)	2.1 (0.5-7.0)	0.7 (0.2-2.1)	1.1 (0.3-2.9)	2.7 (0.9-6.5)	10.9 (4.3-28.2)	< 0.001
HbA1c (%)	6.0 ± 1.1	6.5 ± 1.2	5.8 ± 0.7	5.7 ± 1.0	6.1 ± 1.2	< 0.001
GAP_adm (mmol/L)	0.8 (-0.6-2.8)	-1.4 (-2.5--0.5)	0.3 (-0.4-1.2)	1.6 (0.6-3.0)	4.5 (2.2-7.6)	< 0.001
GAP_mean (mmol/L)	0.9 (-0.4-2.7)	-1.4 (-2.3--0.9)	0.2 (-0.1-0.6)	1.7 (1.3-2.2)	4.7 (3.5-6.5)	< 0.001
Reperfusion therapy
Primary PCI, n (%)	670 (13.5)	107 (8.6)	160 (12.9)	178 (14.4)	225 (18.2)	< 0.001
Thrombolytic therapy, n (%)	2543 (51.4)	589 (47.6)	657 (53.0)	624 (50.5)	673 (54.3)	0.004
Medications
Antiplatelets, n (%)	4812 (97.2)	1190 (96.1)	1203 (97.1)	1214 (98.2)	1205 (97.3)	0.019
Statins, n (%)	3703 (74.8)	938 (75.8)	923 (74.5)	931 (75.3)	911 (73.5)	0.590
β-blockers, n (%)	3171 (64.0)	775 (62.6)	815 (65.8)	785 (63.5)	796 (64.2)	0.405
ACEI/ARB, n (%)	3665 (74.0)	891 (72.0)	932 (75.2)	908 (73.5)	934 (75.4)	0.170
CCB, n (%)	582 (11.8)	175 (14.1)	112 (9.0)	137 (11.1)	158 (12.8)	0.001
Nitrates, n (%)	4596 (92.8)	1149 (95.4)	1157 (93.4)	1138 (92.1)	1120 (90.4)	< 0.001
Diuretics, n (%)	1226 (24.8)	280 (22.6)	269 (21.7)	298 (24.1)	379 (30.6)	< 0.001
Insulin, n (%)	696 (14.1)	105 (8.5)	115 (9.3)	149 (12.1)	327 (26.4)	< 0.001

ACEI angiotensin-converting enzyme inhibitors, ARB angiotensin receptor antagonists, BG blood glucose, CCB calcium channel blockers, CV coefficient of variation, GAP_adm glycemic gap between blood glucose at admission and A1c-derived average glucose, GAP_mean glycemic gap between mean glucose level and A1c-derived average glucose, HbA1c glycosylated hemoglobin, type A1c, LBBB left bundle branch block, MGL mean glucose level, PCI percutaneous coronary intervention, SD standard deviation, STE ST-segment elevation, TIMI the Thrombolysis in Myocardial Infarction

Table 2 Predictors of all-cause mortality and major adverse cardiovascular events by multivariate Cox analysis

Variables	HR represents	All-cause mortality		Major adverse cardiovascular evnets
Variables	HR represents	HR (95% CI)	P value	HR (95% CI)	P value
GAP_mean	Per 1 mmol/L increment	1.10 (1.07-1.14)	< 0.001	1.09 (1.06-1.11)	< 0.001
Age	Per 1 year increment	1.05 (1.04-1.07)	< 0.001	1.03 (1.02-1.04)	< 0.001
Female	Female vs. male	1.33 (1.03-1.72)	0.028	1.13 (0.93-1.37)	0.233
Weight	Per 10 kg increment	0.96 (0.86-1.07)	0.416	0.90 (0.83-0.98)	0.012
Systolic blood pressure	Per 10 mmHg increment	0.91 (0.87-0.94)	< 0.001	0.89 (0.86-0.92)	< 0.001
Heart rate	Per 10 beat/min increment	1.11 (1.06-1.17)	< 0.001	1.06 (1.02-1.10)	0.004
Anterior STE or LBBB	Yes vs. no	1.54 (1.21-1.96)	0.001	1.12 (0.94-1.35)	0.208
Killip classification II-IV	II-IV vs. I	1.18 (1.04-1.34)	0.013	1.39 (1.26-1.53)	< 0.001
Time to treatment > 4 hours	Yes vs. no	1.38 (1.07-1.78)	0.013	1.06 (1.02-1.10)	0.004
Diabetes mellitus	Yes vs. no	1.01 (0.73-1.38)	0.974	1.07 (0.84-1.37)	0.577
Hypertension	Yes vs. no	1.27 (1.00-1.63)	0.052	1.17 (0.97-1.41)	0.094
Myocardial infarction	Yes vs. no	1.00 (0.68-1.47)	0.996	0.96 (0.71-1.29)	0.764
Stroke	Yes vs. no	1.40 (1.03-1.89)	0.032	1.42 (1.12-1.80)	0.004
Heart failure	Yes vs. no	1.25 (0.80-1.98)	0.331	1.01 (0.68-1.49)	0.968
Thrombolysis	Yes vs. no	1.11 (0.87-1.41)	0.401	1.22 (1.01-1.47)	0.035
PCI	Yes vs. no	0.38 (0.22-0.66)	0.001	0.68 (0.49-0.95)	0.025
Antiplatelets	Yes vs. no	0.62 (0.40-0.94)	0.025	0.63 (0.44-0.89)	0.009
Stains	Yes vs. no	0.65 (0.51-0.82)	< 0.001	0.67 (0.56-0.80)	< 0.001
β-blockers	Yes vs. no	0.75 (0.59-0.95)	0.015	0.76 (0.63-0.91)	0.002
ACEI/ARB	Yes vs. no	0.72 (0.56-0.93)	0.013	0.77 (0.64-0.94)	0.009

ACEI angiotensin-converting enzyme inhibitors, ARB angiotensin receptor antagonists, GAP_mean glycemic gap between mean glucose level and A1c-derived average glucose, LBBB left bundle branch block, PCI percutaneous coronary intervention, STE ST-segment elevation

Table 3 NRI and IDI analyses to assess the adding value of GAP_mean in risk stratification^a

	Est.	Lower	Upper	p value
All-cause mortality NRI (Categorical)	0.039	0.003	0.075	*0.033*
NRI (Continuous)	0.216	0.104	0.328	*< 0.001*
IDI	0.013	0.005	0.020	*< 0.001*
Major adverse cardiovascular events
NRI (Categorical)	0.028	-0.002	0.058	*0.063*
NRI (Continuous)	0.198	0.111	0.286	*< 0.001*
IDI	0.010	0.005	0.015	*< 0.001*

GAP_mean glycemic gap between mean glucose level and A1c-derived average glucose, NRI net reclassification improvement, IDI integrated discrimination improvement.

^aThis analysis was performed by comparing the original model (incorporating age, sex, weight, systolic blood pressure, heart rate, anterior ST-segment elevation or left bundle branch block, Killip classification II-IV, time to treatment > 4 hours, diabetes mellitus, hypertension, myocardial infarction, stroke, heart failure, thrombolysis, percutaneous coronary intervention, antiplatelets, statins, β-blockers, angiotensin-converting enzyme inhibitors or angiotensin receptor antagonists) with the new model (incorporating all the variables in the original models and adding GAP_mean)

No competing interests reported.

supplementaryfile1.pdf
Supplementary File 1. Results of receiver operating characteristic curve analysisABG admission blood glucose, GAP_adm defined as the difference between admission blood glucose and A1c-derived average glucose, GAP_mean defined as the difference between mean blood glucose level and A1c-derived average glucose
Supplementaryfile2.pdf
Supplementary File 2. Calibration plot and Hosmer-Lemeshow goodness of fit test of the original model for predicting all-cause mortality and major adverse cardiovascular events

Download PDF

Version 1

posted

You are reading this latest preprint version

Impact of glycemic gap on 30-day adverse outcomes in patients with acute ST-segment elevation myocardial infarction

Status:

Version 1

Abstract

Figures

Backgrounds

Methods

Results

Discussion

Strengths And Limitations

Conclusions

Abbreviations

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1