| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Service d’Endocrinologie (B.V.) and Service de Cardiologie (J.-C.B., J.E.W., Y.C.), Centre Hospitalier Universitaire Bocage, 21034 Dijon, France; Laboratory of Experimental Cardiovascular Pathophysiology and Pharmacology (M.Z.), Institut Fédératif de Recherche 100, Faculties of Medicine and Pharmacy, University of Burgundy, 21000 Dijon, France; Service de Cardiologie (G.D.), Clinique de Fontaine, 21121 Fontaine les Dijon, France; Service de Cardiologie (Y.L.), Centre Hospitalier, 21140 Semur en Auxois, France; Service de Cardiologie (L.J.-M.), Centre Hospitalier, 21200 Beaune, France; and Service de Cardiologie (H.M.), Centre Hospitalier, 21400 Châtillon sur Seine, France
Address all correspondence and requests for reprints to: Bruno Vergès, Service d’Endocrinologie, Centre Hospitalier Universitaire Bocage, Bd Mal de Lattre de Tassigny, 21034 Dijon, France. E-mail: bruno.verges{at}chu-dijon.fr.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Research Design and Methods: A total of 2353 patients with acute MI and surviving at d 5 after admission were analyzed for short-term morbidity and mortality. FG was obtained at d 4 and 5. Patients were classified as diabetes mellitus (known diabetes or FG 
126 mg/dl), high IFG (110 
FG < 126 mg/dl), low IFG (100 FG < 110 mg/dl), and normal fasting glucose (NFG) (FG < 100 mg/dl).
Results: Among the 2353 patients, 968 (41%) had diabetes mellitus, 262 (11%) had high IFG, 332 (14%) had low IFG, and 791 (34%) had NFG. Compared with NFG patients, 30-d cardiovascular mortality was increased in high but not low IFG subjects. In-hospital heart failure was increased in high IFG subjects (42 vs. 20% for NFG, P < 0.0001) but not low IFG subjects (21 vs. 20%). High IFG, but not low IFG, was an independent factor associated with 30-d cardiovascular mortality [odds ratio 2.33 (1.55–3.47)] and in-hospital heart failure [odds ratio 1.70 (1.36–2.07)]. The optimal threshold levels of FG on the receiver-operating characteristic curves were 114 and 112 mg/dl to predict mortality and in-hospital heart failure, respectively.
Conclusion: The present study, based on a nonselected cohort of MI patients, underscores the high prevalence of IFG (25%) and highlights the clinical relevance of 110 mg/dl, but not 100 mg/dl, as a cutoff value to define the risk for worse outcome.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Thus, the aim of our present work was to investigate, using a large nonselected population of MI patients, the impact of glycemic status, stratified by the level of FG, on short-term outcomes after acute MI, including in-hospital heart failure and 30-day cardiovascular (CV) mortality.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The design and methods of the ObseRvatoire des Infarctus de Côte d’Or Survey have been published (3, 6). Briefly, from January 1, 2001, to July 31, 2005, the French regional ObseRvatoire des Infarctus de Côte d’Or survey prospectively collected data from all consecutive patients hospitalized in an intensive care unit for acute MI in all the public centers or privately funded hospitals of one eastern region of France with a population of approximately 500,000 inhabitants. Data were collected at each site by a study coordinator trained in completing the core record form and extracting data from medical records, using a standardized case report form. Cases were ascertained by the prospective collection of consecutive admissions. Internal and external audit checks are performed every year and show that less than 1% of patients might have been missed by the collection procedure. Patients surviving at d 5 after MI onset and presenting with acute MI diagnosed according to European Society of Cardiology and American College of Cardiology criteria were included in the present study (7). MI was defined by an increase in serum troponin I (higher than the upper limit of the hospital normal range) and clinical symptoms of ischemia and/or characteristic electrocardiogram (ECG) signs. ST segment elevation myocardial infarction was diagnosed when new ST segment elevation 1 mm or greater was seen in any location or when new left bundle branch block was found on the qualifying ECG. The present study complied with the Declaration of Helsinki and was approved by the ethics committee of university hospital of Dijon. Informed, written consent from each patient before participation was required for study entry.
Data collection
Data on demographics, CV risk factors, and medical history were collected prospectively along with baseline clinical data and admission data [systolic blood pressure (SBP) and heart rate]. At each participating site, echocardiography was performed at d 3 ± 1 to calculate left ventricular ejection fraction (LVEF). Acute reperfusion procedures [thrombolysis and primary coronary intervention (PCI)] were collected. The short-term outcomes was defined by 30-d mortality as well as in-hospital heart failure, ventricular arrhythmia (ventricular tachycardia or fibrillation), and recurrent MI. Recurrent MI was assessed on ECG modifications and/or increased serum troponin. In-hospital heart failure was defined as rales over more than half of the lung field (Killip class II), pulmonary edema (Killip class III), or cardiogenic shock (Killip class IV). After hospital discharge, 30-d information was acquired by contacting either patients individually or their relatives or treating physician and reviewing the hospital records for cardiac events if the patient had been rehospitalized.
Blood samples were drawn on admission to measure concentrations of glycated hemoglobin (HbA1c) and after overnight fasting (on the day after admission for blood lipids and at d 4 and 5 for determination of FG). FG measured at d 4 and 5 after MI gives reliable estimates of stable glucose metabolism (8). Blood glucose was assessed by the enzymatic method on a Vitros 950 analyzer (Ortho Clinical Diagnostics, Rochester, NY). HbA1c concentration was measured with ion exchange HPLC (Bio-Rad Laboratories, Richmond, CA). Fasting glycemia was calculated by the mean blood glucose values at d 4 and 5. Blood lipid concentrations were determined on a dimension Xpand (Dade Behring, Deerfield, IL) using enzymatic methods. Low-density lipoprotein (LDL) cholesterol was calculated according to the Friedewald formula (9).
In the present study, 100% of the patients were assessed for their glucometabolic state based on complete data for FG at d 4 and 5 and consideration of known diabetes. End points were obtained for all the patients. Among the other variables, CV risk factors, history, clinical data, acute treatments, blood lipids, and HbA1c were more than 95% complete. However, LVEF values were available for only 79% of patients. The proportion of patients with missing data was similar for the four groups (P = 0.896).
Group definition
We classified patients as having DM if they had a history of diagnosed DM or their FG was 126 mg/dl or greater (7 mmol/liter). IFG was defined as mean fasting blood glucose between 100 mg/dl (5.55 mmol/liter) and 126 mg/dl (7 mmol/liter), and patients with IFG were divided into two groups: a high-IFG group, characterized by an FG 110 mg/dl or greater (6.10 mmol/liter) and less than 126 mg/dl (7 mmol/liter) corresponding to the previous ADA IFG definition, and a low-IFG group, characterized by an FG 100 mg/dl or greater (5.55 mmol/liter) and less than 110 mg/dl (6.10 mmol/liter). Patients were classified as having normal fasting glucose (NFG) if their FG was less than 100 mg/dl (5.55 mmol/liter).
Statistical analysis
For continuous variables, a Kolmogorov-Smirnov analysis was performed to test for normality. All continuous data tested were nonnormally distributed and expressed as median and interquartile ranges. Comparisons of continuous variables between two groups of patients were performed by the nonparametric Mann-Whitney U test. Qualitative data expressed as percentages were compared by the 
2 test. The baseline characteristics of the four groups categorized by FG level were compared by the ANOVA for continuous variables and the 2 test for noncontinuous variables.
A stepwise multivariate logistic regression analysis was performed to identify the factors influencing the events (CV mortality, in-hospital heart failure). We entered into a full model the baseline variables with a known relationship with outcomes after MI: age, gender, history of hypertension, prior MI, LDL-cholesterol levels, anterior wall infarction, ST elevation infarction, heart rate at admission, SBP at admission, Killip class at admission, reperfusion therapy, and FG group (NFG, low IFG, high IFG, DM) (10, 11). We used the likelihood ratio test to determine variables to be removed from the model. The accuracy of the model was evaluated by plotting the predicted vs. observed outcomes. Receiver-operating characteristic (ROC) analysis was used to assess the ability of various levels of FG to predict CV mortality and in-hospital heart failure. The ROC curve indicates the probability of a true-positive result as a function of the probability of a false-positive result for all possible threshold values (12). P < 0.05 was considered statistically significant. Statistical analyses were performed using the SPSS software (SPSS, Inc., Chicago, IL).
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
). There was no difference among the four groups with regard to anterior wall infarction and SBP. The comparisons between low IFG, high IFG, and DM along with high IFG and DM are also shown in Table 1.
|
. Thirty-day CV mortality was higher in patients with high IFG (P = 0.001) and DM (P < 0.001) than in those with NFG. The proportions were similar in patients with low IFG and NFG. Thirty-day CV mortality was higher in patients with high IFG than those with low IFG (P = 0.003). In-hospital heart failure was increased in patients with high IFG (P < 0.001) and DM (P < 0.001) when compared with NFG. In-hospital heart failure was similar in patients with low IFG and NFG but higher in patients with high IFG than those with low IFG (P < 0.001) (Table 2).
|
). With regard to in-hospital heart failure, both DM [1.82 (1.50–2.19), P = 0.002] and high IFG [1.70 (1.38–2.08), P = 0.010] were independent factors (Table 3). Gender was not an independent predictor for either CV death or in-hospital heart failure. Low IFG was a predictor of neither 30-d CV mortality nor in-hospital heart failure (Table 3). The accuracy of our model was 95% to predict 30-d CV mortality and 89% to predict in-hospital heart failure.
|
, A and B).
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The diabetic patients in the present study resembles the patients included in Diabetes and Insulin-Glucose Infusion in Acute Myocardial Infarction (DIGAMI) 2 for the baseline characteristics (age, male, and BMI) and HbA1c at admission (13). Our findings, from a contemporary registry, underline the high frequency of abnormal glucose metabolism in the current era of MI because glucose abnormalities (high IFG and DM) were found in more than 50% of patients. Moreover, the proportion of IFG is high, representing 25% of patients with acute MI. This finding is consistent with most recent studies analyzing the frequency of abnormal glucose regulation in coronary artery disease (CAD) patients. In 1612 patients undergoing PCI, abnormalities of FG reached 61%, which was much more than expected (14). The Multinational Euro Heart Survey also pointed out that normal glucose regulation is less common than abnormal glucose regulation in patients with unstable CAD (15). Therefore, our work confirms these findings and extends this concept to patients with acute MI. Knowledge of the glucometabolic state of these patients should influence their management because of the great potential to improve outcome.
An increased risk of short-term mortality and heart failure has been reported in patients with stress hyperglycemia, as defined by high blood glucose on admission (16, 17, 18, 19). However, the use of FG for risk stratification after MI remains infrequent. FG determination after MI is particularly relevant because it shows a stronger association with short-term prognosis after MI than does admission glycemia (11). Furthermore, it has recently been reported, in nondiabetic patients with acute MI, that FG is an independent predictor of abnormal glucose tolerance when admission hyperglycemia is not (20). Norhammar et al. (8) have clearly shown that FG at d 4 or 5 was an independent predictor of abnormal glucose tolerance at 3 months. It is important to note that in our study, at the time of FG assessment (d 4 and 5), no IFG patients were under antidiabetic therapy. Ravid et al. (21) in a small population size study reported a higher unadjusted prevalence of in-hospital mortality in nondiabetic patients with elevated FG. Using a cutoff value of 144 mg/dl for FG, O’Sullivan et al. (22) evidenced high FG as an independent predictor for death after adjustment for age.
A study from our group has shown that IFG, defined as FG 110–126 mg/dl (according to the 1997 ADA definition), was a strong risk factor for developing severe heart failure after MI, even after adjustment for confounding factors (3). Recently Suleiman et al. (11) reported that FG between 110 and 121 mg/dl was an independent factor for 30-d mortality. The ADA recently lowered the cutoff point for IFG from 110 to 100 mg/dl (5). The relevance of this new criterion for IFG to predict the risk for DM and CAD remains controversial (23, 24, 25). Moreover, its prevalence and value for short-term prognosis after MI in patients fulfilling these new criteria were unknown. The present work, on a large unselected population of MI, analyzed patients from the lower range of the new IFG criteria (100–110 mg/dl) and those in the higher range of the new IFG criteria (110–126 mg/dl). The results of our study clearly indicate that only high IFG is associated with adverse events after acute MI, independently of traditional factors influencing the prognosis (3, 11). These findings, further confirmed by ROC curve analysis, also demonstrated the lack of increased risk of the lower range of the new IFG criteria (100–110 mg/dl) in the setting of an acute MI.
Given the high prevalence of IFG in the general population (23, 24, 25) and MI patients (3, 11), determination of short-term prognosis of these patients is an important issue. In the present study, 14% of MI patients had low IFG (100–110 mg/dl), therefore accounting for a substantial proportion of MI population. Thus, the prognosis for this population deserves to be defined. The absence of any association between low IFG (100–110 mg/dl) and in-hospital adverse events and 30-d mortality is an important part of the information provided by this work. These results are confirmed by recent findings that suggest raising the FG cutoff point for predicting 30-d mortality after acute MI to 114 mg/dl (11).
The mechanisms explaining the association between hyperglycemia and adverse outcomes in MI patients are not fully understood. However, experimental studies evidenced the deleterious effects of hyperglycemia on the ischemic myocardium (26). Hyperglycemia has been shown to be directly harmful for cardiomyocytes (27). Hyperglycemia is a reflection of relative insulin deficiency, associated with increased lipolysis and elevated circulating free fatty acids, which may damage cardiac cells (28). Hyperglycemia increases the level of inflammatory markers in MI patients (29). An independent relationship between hyperglycemia and either the no-reflow phenomenon (30) or impaired coronary flow after reperfusion has also been found (31).
Our results demonstrate the existence of an FG threshold, starting at 110 mg/dl, that is a predictor of the short-term risk for death and heart failure after MI. However, in the absence of randomized trials, no data are available on the relevance of glycemic control based on maintaining normoglycemia below the cutoff point of 110 mg/dl in patients with acute MI. In a critical care setting, Van den Berghe et al. (32) demonstrated that maintaining glycemia below the cutoff point of 110 mg/dl by insulin can significantly reduce mortality and morbidity.
Study limitations
Our results were obtained in a Caucasian population and may therefore not apply to other ethnic populations. Moreover, the data reported, in the present study, considered only short-term mortality (30 d mortality) and morbidity (in-hospital events) but did not address long-term outcomes.
In conclusion, our study, based on large nonselected cohort of patients with acute MI, underscores the high prevalence of impaired fasting glucose (25%) and highlights, in the context of IFG, the clinical relevance of 110 mg/dl, not 100 mg/dl, as a cutoff value to define the increased risk for 30-d CV mortality and in-hospital heart failure after acute MI.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Statement: The authors have nothing to disclose.
First Published Online April 10, 2007
Abbreviations: ADA, American Diabetes Association; BMI, body mass index; CAD, coronary artery disease; CV, cardiovascular; DM, diabetes mellitus; ECG, electrocardiogram; FG, fasting glucose; HbA1c, glycated hemoglobin; IFG, impaired FG; LDL, low-density lipoprotein; LVEF, left ventricular ejection fraction; MI, myocardial infarction; NFG, normal FG; PCI, primary coronary intervention; ROC, receiver-operating characteristic; SBP, systolic blood pressure.
Received November 28, 2006.
Accepted March 29, 2007.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
