INTRODUCTION — Patients with diabetes mellitus are at increased risk for myocardial infarction (MI) and diabetes is considered a coronary risk equivalent by the National Cholesterol Education Program [1], since type 2 diabetic patients without a prior MI have the same risk of developing an MI as nondiabetic patients who have already had an MI [2]. (See "Prevalence of and risk factors for coronary heart disease in patients with diabetes mellitus".)
Maintenance of strict glycemic control improves long-term microvascular outcomes in patients with both type 1 and type 2 diabetes. Intensive glycemic control aiming for near-normal levels has been shown to have a substantial beneficial effect (58 percent reduction in major cardiovascular disease [CVD] events) on macrovascular outcomes in type 1 diabetes, when applied early in the disease course and after one to two decades of therapy. In the setting of type 2 diabetes, with several coprevalent risk factors for CVD, any beneficial effects of intensive glycemic control on macrovascular disease are likely to be less than those from similarly stringent control of blood pressure and lipid levels. In the acute setting, the evidence of benefit from strict glycemic control with insulin therapy in patients with acute MI is limited. The evidence in other groups of patient is inconsistent. (See "Glycemic control in critically ill adult and pediatric patients" and "Glycemic control and vascular complications in type 1 diabetes mellitus" and "Glycemic control and vascular complications in type 2 diabetes mellitus".)
Poor glycemic control in diabetic patients and stress hyperglycemia in nondiabetic patients is associated with worse outcomes after acute MI but it is not fully understood as to whether strict glycemic control during acute MI hospitalizations improves outcomes. These issues will be reviewed here.
The possible value of glycemic control in diabetic patients undergoing coronary artery bypass graft surgery is discussed separately. (See "Coronary artery revascularization in stable patients with diabetes mellitus", section on 'Intraoperative and postoperative glycemic control'.)
SERUM VERSUS BLOOD GLUCOSE — Practitioners should be aware that glucose measured in whole blood is generally 12 percent lower than values obtained from serum (or plasma). In the studies cited below, an attempt has been made to specify which was reported.
HYPERGLYCEMIA AND OUTCOME AFTER ACUTE MI — The association between hyperglycemia and outcome after acute MI has been evaluated in two settings: the short-term predictive value of the admission serum glucose in patients with and without diabetes and the long-term increase in risk in patients with diabetes.
Predictive value of admission glucose — There is a positive association between the serum glucose at the time of MI and mortality in patients with and without diabetes [3-8]. The prognostic significance of what has been presumed to represent stress hyperglycemia was addressed in an analysis of 15 trials that reported in-hospital mortality after an MI in relation to admission serum glucose concentrations [3]. The analysis comes to the following conclusions of these trials in which the admission glucose were random and may have represented fasting or postprandial states:
●In patients without diabetes, those with glucose concentrations between 110 and 143 mg/dL (6.1 to 8 mmol/L) had a 3.9-fold higher risk of death compared with patients with lower glucose concentrations. Glucose values between 144 and 180 mg/dL (8 to 10 mmol/L) were associated with a three-fold higher risk of heart failure or cardiogenic shock.
●Diabetic patients with glucose concentrations ≥180 to 196 mg/dL (10 to 11 mmol/L) also had an increased risk of death compared with normoglycemic diabetic patients (relative risk 1.7), but this relative risk was lower than in non-diabetics.
In addition to the admission value, a graded relationship has been found between the fasting glucose obtained within 24 hours of admission and 30-day mortality [9,10].
The possible mechanisms by which stress hyperglycemia is associated with worse outcomes were evaluated in a review of 460 consecutive patients with STEMI who were treated with primary percutaneous coronary intervention (PCI); 322 (70 percent) had a serum glucose ≥140 mg/dL (7.8 mmol/L) on admission, but only 14 percent had a history of diabetes [11]. The patients with hyperglycemia were significantly less likely to have TIMI grade 3 (normal) flow before PCI compared with those with normoglycemia (12 versus 28 percent, adjusted odds ratio [OR] 2.6 for the absence of reperfusion). The impairment in coronary flow might reflect a prothrombotic state or endothelial dysfunction associated with hyperglycemia, more severe disease leading to a greater stress response, or hyperglycemia might be a marker for some other determinant of outcome.
Despite the observation of worse outcomes in patients with higher glucose concentrations on admission and within 24 hours, the available studies do not rule out the possibility that high blood glucose in this setting is a marker for a sicker patient, as opposed to being causative for worse outcomes.
J- or U-shaped curve — A review of 4224 patients in trials of fibrinolysis or primary PCI in patients with STEMI found a U-shaped relationship between the serum glucose (mostly admission values) and the 30-day rate of death or recurrent MI [12]. The following findings were noted in patients with hypoglycemia, defined as blood glucose values of <81 mg/dL (4.5 mmol/L), euglycemia, defined as blood glucose values of 81 to 99 mg/dL (4.5 to 5.5 mmol/L), and severe hyperglycemia, defined as blood glucose values >199 mg/dL (11.0 mmol/L).
●The 30-day mortality rate in the three groups was 4.6, 1.0, and 4.7 percent, respectively
●The 30-day rate of recurrent MI or death in the three groups was 10.5, 4.2, and 7.2 percent, respectively
After adjustment for baseline differences, the relative risk for mortality was significantly increased, compared with the euglycemic group, in patients with hypoglycemia (OR 3.37) or severe hyperglycemia (OR 3.09). The risk was also significantly increased in patients with blood glucose values between 150 and 199 mg/dL (8.3 and 11.0 mmol/L, OR 2.93). The U-shaped relationship was seen in both diabetic and nondiabetic patients.
Similar findings (higher mortality with both hyperglycemia on admission and hypoglycemia during hospitalization) were noted at two years in a review of 713 consecutive patients with diabetes and non-ST elevation acute coronary syndrome [13]. Both persistent hypoglycemia and hyperglycemia were found to be better predictors of mortality than admission glucose in a separate study [14].
Other observational studies have also demonstrated a positive relationship between admission or in-hospital hyperglycemia, as well as glycemic variability, with a variety of complications following acute MI [15-17].
It is not yet clear from these data whether treatment-induced hypoglycemia is driving increased risk. One analysis of hyperglycemia in patients with acute MI found that spontaneous hypoglycemia was associated with increased hospital mortality, but not hypoglycemia associated with insulin therapy [18].
Conceivably, other medical factors, such as malnutrition, hepatic or renal disease or sepsis may be playing a role in these observational data.
Worse outcomes in diabetic patients — The long-term outcome after an acute MI is worse in diabetic patients than nondiabetic patients with both a non-ST elevation MI (NSTEMI) and an ST elevation MI (STEMI). The adverse effect is manifested by increases in mortality and nonfatal cardiovascular end points (such as reinfarction or heart failure) [2,19]. Diabetic patients tend be older and to have a greater prevalence of comorbidities compared with patients without diabetes [19]. However, the increase in risk persists after adjustment for these differences. The data supporting these conclusions is presented separately.
UNDIAGNOSED DIABETES — The frequency with which patients with an acute MI have undiagnosed diabetes was addressed in a prospective study in which glucose metabolism (via fasting blood glucose and oral glucose tolerance test) was evaluated in 181 consecutive patients with an acute MI and no prior history of diabetes [20]. The criteria for diagnosing diabetes were a fasting plasma glucose ≥126 mg/dL (7 mmol/L) and/or two-hour postload plasma glucose above 200 mg/dL (11.1 mmol/L). The respective values for impaired glucose tolerance were a fasting plasma glucose between 110 and 125 mg/dL (6.1 to 6.9 mmol/L) and a two hour postload plasma glucose between 140 and 200 mg/dL (7.8 to 11.1 mmol/L).
The following findings were noted:
●Impaired glucose tolerance was present in 35 percent at hospital discharge and 40 percent three months later. The respective values for previously undiagnosed diabetes were 31 and 25 percent at these time periods. However, the incidence of previously undiagnosed diabetes was much lower (10 and 13 percent) when only the fasting blood glucose was used, as suggested by the American Diabetes Association. Whether the same patients had abnormal glucose metabolism at discharge and three months is not clear. (See "Clinical presentation, diagnosis, and initial evaluation of diabetes mellitus in adults".)
●Independent predictors for abnormal glucose tolerance at three months were the hemoglobin A1C concentration on admission and the fasting blood glucose on day four. The former observation, which indicates that the metabolic abnormality preceded the infarction, and the persistence of abnormal glucose tolerance at three months suggest that the hyperglycemia during hospitalization cannot be simply explained as stress hyperglycemia in some patients.
The presence of elevated levels of Hb A1c found in most of these patients makes stress hyperglycemia a less likely diagnosis. At the least, these findings strongly suggest that the fasting plasma glucose concentration and hemoglobin A1c should be measured during hospitalization in nondiabetic patients with an acute MI and that elevated values be repeated after discharge to identify those at increased risk.
VALUE OF GLYCEMIC CONTROL — Evidence from which recommendations for glycemic control in patients with acute MI comes from studies in three populations: critically ill patients (a minority of whom have an acute MI) in intensive care units; patients admitted with acute MI; patients admitted to general medical wards.
Critically ill patients — Randomized trials have assessed the efficacy of strict glycemic control (target blood glucose 80 to 110 mg/dL [4.4 to 6.1 mmol/L]) compared with standard care (180 to 200 mg/dL; 10.0 to 11.1 mmol/L) in patients admitted to an intensive care unit (ICU). The studies evaluating the efficacy of strict glycemic control in patients admitted to ICUs and recommendations for glycemic control in such patients are discussed elsewhere. (See "Glycemic control in critically ill adult and pediatric patients".)
Although initial single-center studies suggested a strong benefit, mainly in post-surgical patients, these data have not been replicated by others. The largest of the randomized trials in ICU patients, the NICE-SUGAR trial, found an increase in the rate of death at 90 days with intensive glucose control (81 to 108 mg/dL; 4.5 to 6.0 mmol/L) as compared with those whose blood glucose was maintained between 144 and 180 mg/dL (8.0 to 10.0 mmol/L) [18]. In addition, the data from these studies may or may not apply to patients with MI. The optimal blood glucose target in hyperglycemic critically ill patients is uncertain, but we believe the compromise of 140 to 180 mg/dL (7.8 to 10.0 mmol/L) seems reasonable.
Patients with acute MI who might be labeled critically ill include those with hemodynamic or electrical instability, acute heart failure, or coexistent serious acute illnesses such as acute exacerbation of obstructive lung disease requiring respiratory support or acute blood loss requiring transfusion.
Studies of patients with acute myocardial infarction — A number of trials have attempted to address the value of intensive insulin therapy and/or glycemic control specifically in patients with an acute MI. None was as large and well designed as the aforementioned ICU trials and each has important flaws. Furthermore, none of the trials achieved the level of glucose control that would have been considered intensive in the ICU trials.
DIGAMI trial — In the DIGAMI trial, 620 diabetic patients with an acute MI were randomly assigned to an insulin-glucose infusion for 24 hours followed by subcutaneous insulin four times daily for ≥3 months or standard treatment with insulin therapy only if clinically indicated [21]. The target blood glucose level for patients assigned to the insulin-glucose infusion was 126 to 196 mg/dL (7 to 10.9 mmol/L).
With respect to glycemic control, the following findings were noted:
●At randomization, the mean blood glucose was about 280 mg/dL (15.6 mmol/L). The blood glucose was significantly lower with intensive insulin at 24 hours (173 versus 211 mg/dL [9.6 versus 11.7 mmol/L]) and hospital discharge (148 versus 162 mg/dL [8.2 versus 9.0 mmol/L]).
●At randomization, the HbA1c was 8.1 percent. The reduction in HbA1c was significantly greater with intensive insulin therapy at three months (1.1 versus 0.4 percent) and one year (0.9 versus 0.4 percent).
Mortality was significantly lower in the group assigned to more aggressive insulin therapy at one year (19 versus 26 percent) and at 3.4 years (33 versus 44 percent) [4,21]. The greatest reduction in mortality was seen in low-risk patients who had not been receiving insulin prior to the infarction.
Since DIGAMI also included an outpatient insulin therapy component, the isolated effect of glycemic control in-hospital could therefore not be easily assessed.
DIGAMI-2 trial — The value of insulin therapy was further addressed in the DIGAMI-2 trial, in which patients with type 2 diabetes and acute MI were randomly assigned to one of three glucose management strategies: group 1, inpatient insulin infusion/outpatient intensive subcutaneous insulin therapy; group 2, inpatient insulin infusion/outpatient standard treatment; or group 3, inpatient/outpatient routine glucose management according to local practice [22].
Although it was anticipated that mortality rates would be lowest in group 1, they were similar in all three groups. However, there were a number of problems with this study that interfere with interpretation of the results:
●Glycemic control, which was expected to be the best in group 1, was also similar in the three groups.
●The overall event rate was lower than expected in all groups (perhaps due to implementation of other secondary prevention strategies), which may have attenuated any statistical differences between groups.
●The trial was stopped earlier than planned due to a failure to recruit an adequate number of patients; since less than 50 percent of the required patients were recruited, the power to detect a difference among the treatment groups was substantially reduced.
HI-5 trial — The possible benefit of more intensive glucose control in patients with an acute MI with either a history of diabetes or an admission blood glucose ≥140 mg/dL (7.8 mmol/L) was evaluated in the Hyperglycemia: Intensive Insulin Infusion in Infarction (HI-5) study [23]. In this trial, 240 such patients were randomly assigned to conventional therapy or to an insulin/dextrose infusion to maintain the blood glucose between 72 and 180 mg/dL (4 and 10 mmol/L) for at least 24 hours. After 24 hours, the patients were managed with standard care by their own physicians with a recommend HbA1c of less than 7 percent.
There was no difference in the primary end point of mortality in-hospital or at three or six months. However, HI-5 was seriously flawed by the small number of patients, lack of blinding, maintenance of glycemic control for only 24 hours, and failure to attain a significant difference in mean 24-hour blood glucose between the intensive therapy and control groups (149 versus 162 mg/dL [8.3 versus 9.0 mmol/L]).
Subset analysis found that patients who had a mean blood glucose ≤144 mg/dL (8.0 mmol/L) during the first 24 hours had large reductions in mortality in-hospital (0 versus 7 percent) and at three and six months (2 versus 11 percent). The possible mortality benefit in this subgroup analysis is consistent with the above surgical ICU trial but needs to be confirmed in patients with an acute MI.
Observational evidence — Additional evidence supporting a benefit from glycemic control comes from an observational study of 7820 hyperglycemic patients (admission blood glucose ≥140 mg/dL; 7.8 mmol/L) hospitalized with acute MI [24]. After multivariable adjustment, lower mean postadmission glucose levels were associated with better all-cause in-hospital mortality. There was no difference in mortality rates between insulin-treated and non-insulin treated patients irrespective of mean postadmission glucose level.
General medical patients — Most patients hospitalized with acute MI are not critically ill. Indeed, with early reperfusion using either percutaneous coronary intervention or fibrinolytic therapy, most are not classified as critically ill and many are discharged within 48 to 72 hours. Therefore, it may not be appropriate to apply evidence from studies of critically ill patients, or even older studies of MI.
However, there is limited evidence, from studies of general medical patients regarding the optimal management of hyperglycemia, which can be applied to patients with acute MI. That evidence suggests that reasonable glycemic goals are to avoid hypoglycemia (fasting blood glucose concentrations no lower than 90 mg/dL [5.0 mmol/L]) and to keep all glucose levels below 200 mg/dL (11.1 mmol/L) in an attempt to avoid dehydration, caloric loss, glycosuria, and to reduce the risk of infection and, although rare, ketoacidosis. This issue is discussed in detail separately. (See "Management of diabetes mellitus in hospitalized patients".)
Summary — The optimal strategy for the management of hyperglycemia in patients with and without diabetes admitted with acute MI is unknown. Studies of critically ill patients may not be relevant, studies in patients with acute MI are flawed, and evidence from studies of patients admitted to general medical wards is limited.
We believe that there is general consensus around the following points:
●Glucose control should be initiated when serum glucose values are above 180 to 200 mg/dL (10 to 11 mmol/L).
●Hypoglycemia, defined as a fasting blood glucose concentration below 70 mg/dL [(3.9 mmol/L]) should be avoided. To ensure safety, a reasonable lower goal might be to maintain all blood glucose levels >90 to 100 mg/dL (5 to 5.6mmol/L).
●Critically ill patients, such as those in cardiogenic shock, represent as small subset of all those with an acute MI. A recent trend is to recommend that blood glucose be maintained between 140 and 180 mg/dL (7.8 to 10 mmol/L) with intravenous insulin in these individuals in the ICU [3,25-27].
METHOD OF TREATMENT — The discussions of how to achieve glycemic control in critically ill patients and in patients with diabetes admitted to general medical wards are found elsewhere. (See "Management of diabetes mellitus in hospitalized patients", section on 'Prevention and treatment of hyperglycemia' and "Glycemic control in critically ill adult and pediatric patients", section on 'Our approach'.)
RECOMMENDATIONS OF OTHERS — We agree with the 2014 American College of Cardiology/American Heart Association guideline for the management of patients with non-ST-elevation acute coronary syndromes, which recommends achieving and maintaining blood glucose less than 180 mg/dL (10 mmol/L) [28]. During this time, hypoglycemia should be avoided. This recommendation was made for patients with and without a diagnosis of diabetes. This guideline does not make a recommendation regarding long-term glucose control. This issue is discussed separately. (See "Initial management of hyperglycemia in adults with type 2 diabetes mellitus" and "Glycemic control and vascular complications in type 2 diabetes mellitus", section on 'Macrovascular disease'.)
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Non-ST-elevation acute coronary syndromes (non-ST-elevation myocardial infarction)" and "Society guideline links: ST-elevation myocardial infarction (STEMI)".)
SUMMARY AND RECOMMENDATIONS — The correction and prevention of hyperglycemia have become standard care for hospitalized patients, including those with acute myocardial infarction (MI). However, whether control of hyperglycemia is sufficient to reduce morbidity and mortality is not proven at this time. The evidence upon which recommendations for glycemic control can be made in such patients is weak for the following reasons:
●The randomized trials in patients with MI are significantly flawed.
●The best data supporting glycemic control come from the trials of critically ill patients in intensive care units, but these trials included few patients with MI. In addition the conclusions of some studies differ from others.
●Most patients admitted with acute MI are not critically ill.
While there is general agreement that glucose value above 200 mg/dL (11 mmol/L) should be treated, there is insufficient evidence to establish an acceptable, minimal blood glucose (treatment target). Although lowering of high blood glucose levels may decrease the risk of poor clinical outcomes, overtreatment leading to hypoglycemia is associated with poor outcomes and hypoglycemia should be strictly avoided. (See 'Value of glycemic control' above and 'J- or U-shaped curve' above.)
For both stable and unstable patients with acute MI with hyperglycemia, including patients with and without diabetes, we suggest an insulin based regimen to achieve and maintain blood glucose less than 180 mg/dL (10 mmol/L) (Grade 2B). There is insufficient evidence to establish a minimal acceptable blood glucose, but the avoidance of hypoglycemia is logical.
The discussions of how to achieve glycemic control in critically ill patients and in patients with diabetes admitted to general medical wards are found elsewhere. (See "Management of diabetes mellitus in hospitalized patients", section on 'Prevention and treatment of hyperglycemia' and "Glycemic control in critically ill adult and pediatric patients", section on 'Our approach'.)
More data are needed to inform a decision as to whether a more stringent target is justified or even safe, based on the trends seen in clinical trials of other critically ill patients.
