AJCC
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

American Journal of Critical Care. 2002;11: 504-519
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Respond to This Article
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Right arrow Take the CE Test
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chyun, D.
Right arrow Articles by Krumholz, H. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chyun, D.
Right arrow Articles by Krumholz, H. M.

Cardiac Outcomes After Myocardial Infarction in Elderly Patients With Diabetes Mellitus

By Deborah Chyun, RN, PhD, Viola Vaccarino, MD, PhD, Jaime Murillo, MD, Lawrence H. Young, MD and Harlan M. Krumholz, MD. From Yale University, New Haven, Conn.


   Abstract
Top
 Abstract
Methods
 Results
 Discussion
 Clinical Implications
 Limitations
 Conclusions
 References
 
• Objective To examine the association between (1) comorbid conditions related to diabetes mellitus, clinical findings on arrival at the hospital, and characteristics of the myocardial infarction and (2) risk of heart failure, recurrent myocardial infarction, and mortality in the year after myocardial infarction in elderly 30-day survivors of myocardial infarction who had non–insulin- or insulin-treated diabetes.

• Methods Medical records for June 1, 1992, through February 28, 1993, of Medicare beneficiaries (n = 1698), 65 years or older, hospitalized for acute myocardial infarction in Connecticut were reviewed by trained abstractors.

• Results One year after myocardial infarction, elderly patients with non–insulin- and insulin-treated diabetes mellitus had significantly greater risk for readmission for heart failure and recurrent myocardial infarction than did patients without diabetes mellitus, and risk was greater in patients treated with insulin than in patients not treated with insulin. Diabetes mellitus, comorbid conditions related to diabetes mellitus, clinical findings on arrival, and characteristics of the myocardial infarction, specifically measures of ventricular function, were important predictors of these outcomes. Mortality was greater in patients not treated with insulin than in patients treated with insulin; the increased risk was mostly due to comorbid conditions related to diabetes mellitus and poorer ventricular function.

• Conclusions Risk of heart failure, recurrent myocardial infarction, and mortality is elevated in elderly patients who have non–insulin- or insulin-treated diabetes mellitus. Comorbid conditions related to diabetes mellitus and ventricular function at the time of the index myocardial infarction are important contributors to poorer outcomes in patients with diabetes mellitus.

After acute myocardial infarction, diabetes mellitus is associated with an increased risk of heart failure,1,2 recurrent myocardial infarction,1,3–7 and mortality.1–5,7–26 Although diabetes mellitus and myocardial infarction are important problems in elderly patients, with an 18% to 20% prevalence of diabetes mellitus in adults 60 to 74 years old,27,28 few elderly patients have been studied after myocardial infarction. In addition, most of the previous investigations were done before the thrombolytic era,1,3–5,8–10,13–19 so other than results based on databases from clinical trials,2,6,7,20–22 current data comparing subjects with and without diabetes mellitus are limited.11,12,23,25 In thrombolytic clinical trials, patients were excluded from participation if they had heart failure, shock, non–Q-wave myocardial infarction, or contraindications to thrombolytic therapy. Thus, the findings from these trials may not be applicable to the larger population of patients with myocardial infarction, particularly elderly patients.

In addition, although the importance of heart failure and myocardial infarction as prognostic factors for mortality is well documented in patients with myocardial infarction who have diabetes mellitus,1,2,14,29 factors predictive of the development of heart failure and recurrent myocardial infarction in the period after infarction have not been determined. The occurrence of comorbid conditions, which are more prevalent in elderly patients with diabetes mellitus than in similar patients without diabetes mellitus, has not been assessed in relation to these outcomes, and the reasons for the elevated risk in subjects with diabetes mellitus have not been explained. Knowledge of the characteristics of myocardial infarction and factors associated with a poor long-term prognosis would assist not only in anticipating potential problems in the period after acute myocardial infarction but also in determining which patients are at high risk and could serve as a basis for discharge planning.

The purpose of our investigation was to examine the association between (1) comorbid conditions related to diabetes mellitus, clinical findings on arrival at the hospital, and characteristics of the myocardial infarction and (2) the risk for heart failure, recurrent myocardial infarction, and mortality in the year after myocardial infarction in elderly 30-day survivors of myocardial infarction who had non–insulin- and insulin-treated diabetes mellitus (NIRxDM and IRxDM patients, respectively). We hypothesized that risks for heart failure, recurrent myocardial infarction, and mortality during the year after myocardial infarction are higher in IRxDM patients than in NIRxDM patients or patients without diabetes mellitus and that the elevated risk in patients with NIRxDM and IRxDM is due to differences in comorbid conditions related to diabetes mellitus, clinical findings on arrival at the hospital, and characteristics of the myocardial infarction.


   Methods
Top
 Abstract
 Methods
Results
 Discussion
 Clinical Implications
 Limitations
 Conclusions
 References
 
Study Design and Data Collection
The medical records of all Medicare beneficiaries discharged from 35 acute care hospitals in Connecticut with a principal diagnosis of acute myocardial infarction, identified through International Classification of Diseases, version 9, coding (ICD-9-410), during the 9-month period June 1, 1992, through February 28, 1993, were reviewed. Data were obtained from patients’ records previously examined in the Cooperative Cardiovascular Project by trained medical records abstractors.30 The Cooperative Cardiovascular Project was conducted by the Health Care Financing Administration (now the Centers for Medicare and Medicaid Services) through the Connecticut Peer Review Organization. The purpose of the project was to profile hospital care received by Medicare patients with acute myocardial infarction in order to provide feedback to hospitals and physicians about adequacy of care. Connecticut was 1 of 4 pilot states. These records were reabstracted for our study, which was specifically designed to supplement the data obtained in the original project with more detailed information on patients’ history, including diabetes mellitus status, comorbid conditions related to diabetes mellitus, characteristics of the myocardial infarction, left ventricular ejection fraction (LVEF), complications associated with myocardial infarction, and medications in use at the time of discharge.

Sample
The records of all patients 65 years or older with myocardial infarction at the time of admission, as verified by elevations in serum levels of creatine kinase–MB (CK-MB) greater than the upper limit of normal for each institution within 48 hours of admission, were included in our study. Troponin levels were not widely used at the time of the Cooperative Cardiovascular Project, and these data were not available in the medical records. Clinical history and/or electrocardiographic changes suggestive of myocardial infarction (ST-segment elevation and/or development of new Q waves) were used to validate myocardial infarction in patients who died before blood samples were obtained for measurement of CK-MB levels or before elevation in the enzyme would have been detected. Records of patients transferred from another institution were excluded because information on the patients’ initial clinical findings may not have been available. The original Cooperative Cardiovascular Project had a total of 3747 patients; of these, 2468 met the inclusion criteria for our study. We additionally excluded the records of 24 patients for whom the diagnosis of diabetes mellitus could not be determined. An additional 394 records were for a period outside the 9-month period.

The initial sample, therefore, consisted of 2050 patients. At 30 days after myocardial infarction, 352 patients had died: 17% of the nondiabetic patients, 19% of the NIRxDM patients, and 18% of the IRxDM patients (P = .44). Because the focus of our study was long-term outcomes in survivors of myocardial infarction, the records of these 352 patients were excluded from the analysis, yielding the final study population of 1698 patients. In both projects, reliability for all variables was greater than 90%; this value was formally quantified via reabstraction of a 5% random sample of the records.

Definitions
History of diabetes mellitus was ascertained by evidence in the medical record that the patient had a history of this disease before admission. Information available in the medical record did not allow differentiation between type 1 and type 2 diabetes. Among IRxDM patients, 5% had type 1 diabetes mellitus, and 21% had type 2; the type was not specified for the remaining 74%. Therefore diabetes was classified as NIRxDM and IRxDM. Classification was based on treatment before admission rather than on treatment at the time of discharge.

NIRxDM was chosen if 1 or more of the following terms were used to describe the diabetes: non–insulin-dependent diabetes mellitus; type 2 (or II) diabetes; maturity-onset diabetes mellitus; ketosis-resistant diabetes; stable diabetes; diet-controlled diabetes; or adult-onset diabetes mellitus, with or without use of antidiabetic oral agents or diet management. Any patient who was taking an antidiabetic oral agent or was on a diet for patients with diabetes, even if the record did not specify one of the NIRxDM terms listed here, was classified as NIRxDM.

IRxDM was chosen if 1 or more of the following terms were used: insulin-dependent diabetes mellitus, juvenile diabetes, juvenile-type diabetes, juvenile-onset diabetes mellitus, ketosis-prone diabetes, brittle diabetes, or type 1 (or I) diabetes. Any patient who was taking insulin before admission was classified as IRxDM, even if the IRxDM terms were not used in the patient’s medical record.

Data on other variables used in the analysis were abstracted from the medical records according to whether the variable was specified or not. If a variable was not mentioned in a patient’s record, we assumed that the variable did not occur. These variables included demographic characteristics (age, sex, and ethnicity), comorbid conditions related to diabetes mellitus (history of peripheral vascular disease [PVD] and chronic renal insufficiency [CRI]), clinical findings on arrival at the hospital (functional and mental status), and characteristics of the myocardial infarction (complications after the infarction, LVEF, type of myocardial infarction, laboratory values within 24 hours of admission, and medications prescribed at discharge).

Assessment of heart failure at the time of admission or later during the hospitalization was based on the timing of the episode and on data indicating that the patient had rales; mild, moderate, or severe shortness of breath; or radiological confirmation of heart failure or pulmonary edema. Electrocardiographic data were interpreted by doing a parallel chart abstraction. Other diabetes-related comorbid conditions (previous heart failure, stroke, myocardial infarction, hypertension, and coronary artery bypass surgery) were obtained from the original Cooperative Cardiovascular Project database.

CK-MB levels were measured on a variety of different scales within the different hospitals. Ratios of peak CK levels were therefore calculated by dividing the actual value by the value for the upper limit of normal for the relevant institution, and these ratios were categorized into levels on the basis of the quartiles observed in our study sample (<1.5, 1.5–2.9, 3.0–6.1, and >6.1 times the upper limit of normal). In addition, because some patients (1%) did not have a CK level recorded in their charts, a category was added for those with missing data on CK levels.

LVEF during hospitalization was categorized into 4 levels according to either the actual LVEF value or, if this value was not provided, according to verbal descriptors: level 1, normal or mild depression or an LVEF greater than 0.40; level 2, mild-moderate or moderate depression or an LVEF between 0.30 and 0.40; level 3, moderate-severe or severe depression or an LVEF less than 0.30; and level 4, no LVEF determined.

Assessment of Outcomes
Occurrence and timing of heart failure and myocardial infarction were obtained by using International Classification of Diseases, version 9, codes for hospitalization for these 2 outcomes and the Medicare Enrollment Database. Time of death was also obtained. In analyses of readmission for heart failure and recurrent myocardial infarction, patients who died were included until the time of death.

Statistical Analysis
Preliminary bivariate analyses with {chi}2 tests, analysis of variance, or t tests (SAS Statistical Software 6.11, SAS, Cary, NC) were used to examine the relationships between (1) NIRxDM and IRxDM and potential prognostic factors and (2) between each outcome (ie, heart failure, recurrent myocardial infarction, and mortality) and potential prognostic factors. The Mantel-Haenszel {chi}2 statistic, which is used to test for a linear association between row and column variables when both variables lie on an ordinal scale, was used to provide an estimate of the relationship of each prognostic factor to diabetes status.31 Thus, in addition to the Pearson {chi}2 statistic, the Mantel-Haenszel {chi}2 statistic provided evidence that the association was graded from patients without diabetes mellitus to those with NIRxDM and IRxDM. The relationships of NIRxDM and IRxDM to heart failure, recurrent myocardial infarction, and mortality were then examined by using life-table methods, and the unadjusted survival experience for nondiabetic, NIRxDM, and IRxDM patients was compared by using the log-rank test.32

After the assumption of proportionality was tested and confirmed, Cox proportional hazards regression was used to test the study hypotheses, with controlling for prognostic indicators.33 Continuous variables were entered into the model as a categorical variable based on quartiles if a departure from a linear trend occurred; otherwise they were retained as a continuous variable in the model.34 The variables NIRxDM and IRxDM were included first to obtain unadjusted estimates of the risk. We theorized that the risks of long-term heart failure, myocardial infarction, and mortality are influenced by comorbid conditions related to diabetes mellitus, including underlying cardiac function, and by characteristics of the myocardial infarction, present both at arrival and during the hospitalization. On the basis of clinical judgment and variables associated with each outcome in bivariate comparisons (P < .05), these variables were entered by using a forward-selection strategy, with groups of variables representing comorbid conditions, clinical findings on arrival, and characteristics of the myocardial infarction added in this specified order. Variables were retained at each step if the P value associated with the Wald {chi}2 statistic was .05 or less, if the change in model log-likelihood exceeded the {chi}2 critical value associated with the model degrees of freedom, or if removal resulted in a 10% or greater change in the parameter estimates of other variables in the model.

Model selection was based on minimizing the deviance. Variables removed in earlier steps were then added to the final model to determine if the added variables were important in the presence of other variables; the added variables were removed if they did not meet the specified criteria. Interactions were tested between NIRxDM and IRxDM and each of the variables retained in the final model.

In order to better understand the impact of the identified prognostic factors on the risk for each outcome in patients with NIRxDM and IRxDM, the change in risk estimates for NIRxDM and IRxDM patients was assessed as each variable was added to the model 1 variable at a time. This method provided an assessment of the contribution of each of the variables in explaining the risk for NIRxDM and IRxDM patients. In addition to P values, 95% CIs were calculated to provide a measure of the accuracy of the risk estimates.


   Results
Top
 Abstract
 Methods
 Results
Discussion
 Clinical Implications
 Limitations
 Conclusions
 References
 
Characteristics of the Sample
A total of 476 patients (28%) had a history of diabetes; 298 (63%) were classified as having NIRxDM and 178 (37%) as having IRxDM. Characteristics of patients with and without diabetes mellitus are shown in Table 1Go. The percentage of women was slightly greater for patients with IRxDM than for patients with NIRxDM or patients without diabetes mellitus, although the difference was not significant. Ethnic differences were also apparent; the percentages of patients who were not white were greater for patients with IRxDM than for the other 2 groups. Patients with IRxDM were significantly younger than those with NIRxDM or without diabetes mellitus.


View this table:
[in this window]
[in a new window]
  		Table 1 Characteristics of subjects without diabetes mellitus and with non–insulin-treated and insulin-treated diabetes mellitus (n = 1698)*

 
Compared with patients without diabetes mellitus, patients with NIRxDM and IRxDM were more likely to have had previous heart failure, stroke, CRI, myocardial infarction, PVD, hypertension, and coronary artery bypass graft surgery and were more likely to have needed assistance in 1 or more activities of daily living before admission to the hospital. Compared with patients without diabetes mellitus, at the time of admission, patients with NIRxDM and IRxDM were more likely to have had signs and symptoms of heart failure, higher body mass index, and higher heart rate and systolic blood pressure. Patients with IRxDM were the least likely to have had chest pain. Prevalence of anterior myocardial infarction was similar in patients with and without diabetes mellitus. Heart failure after the first 24 hours, as either a new event or as a recurrence in patients who had heart failure initially, occurred with almost equal frequency in patients with and without diabetes mellitus. Differences were not statistically significant for presence of Q-wave versus non–Q-wave myocardial infarction or for LVEF. However, patients with NIRxDM and IRxDM tended to have a lower LVEF than did patients without diabetes mellitus (Mantel-Haenszel P=.05).

Patients with diabetes were more likely than patients without the disease to have CK ratios 1.5 times or less than the normal limit and less likely to have ratios more than 6.1 times the normal limit. Compared with patients without diabetes, patients with NIRxDM and IRxDM were less likely to have received thrombolytic agents; however, they were also less likely to have had ST-segment elevation (data not given) at the time of arrival. The frequency of administration of aspirin upon arrival and use of intravenous heparin was equal in all 3 groups of patients (data not given). Potassium levels within the first 24 hours, categorized into quartiles, were higher in patients with NIRxDM and IRxDM than in patients without diabetes.

Patients with NIRxDM and IRxDM were more likely than patients without diabetes to be taking digoxin, diuretics, angiotensin-converting enzyme inhibitors, and calcium antagonists at the time of discharge, and they tended to be taking nitrates (Mantel Haenszel P = .05). Although differences were not observed between the 3 groups for use of aspirin or ßblockers at the time of discharge, most of the patients were not taking these medications at that time.

Readmission for Heart Failure
At 1 year after the myocardial infarction, 230 patients had been admitted to the hospital again because of heart failure: 135 (11%) of the patients without diabetes mellitus, 50 (17%) of the patients with NIRxDM, and 45 (25%) of the patients with IRxDM (P = .001). The risk for readmission because of heart failure among the 3 groups during the year after myocardial infarction is shown in Figure 1Go. Of the 3 groups, patients with IRxDM were most likely to have a readmission; differences among the 3 groups were apparent soon after the myocardial infarction (P < .001).



View larger version (8K):
[in this window]
[in a new window]
  		Figure 1 Cumulative rates of readmission for heart failure.

 
In order to obtain an estimate of the risk in patients with NIRxDM and IRxDM compared with patients without diabetes, Cox proportional hazards regression was used. Unadjusted relative risk estimates were 1.59 (95% CI = 1.15–2.20) for patients with NIRxDM and 2.55 (95% CI = 1.82–3.57) for patients with IRxDM. Thus, compared with patients without diabetes, patients with NIRxDM were 1.59 times more likely and patients with IRxDM were 2.55 times more likely to have a readmission for heart failure. After age was controlled for, compared with the risk for patients without diabetes, the relative risk was 1.63 (95% CI = 1.18–2.25) for patients with NIRxDM and 2.99 (95% CI = 2.12–4.21) for patients with IRxDM.

With adjustments for comorbid conditions related to diabetes mellitus, clinical findings on arrival at the hospital, and characteristics of the myocardial infarction in the multivariate model (Table 2Go), risk of heart failure in patients with NIRxDM (RR = 1.19; 95% CI = 0.85–1.67) and IRxDM (RR = 2.00; 95% CI = 1.40–2.86) dramatically decreased. Increasing age, previous coronary artery bypass graft surgery and CRI, comatose mental status, heart failure, and increased heart rate on arrival at the hospital, lower LVEF, presence of heart failure during hospitalization, and treatment with diuretics at the time of discharge were associated with an increased risk of heart failure in the following year. In additional analyses, when a forward approach to model building was used, heart failure and heart rate at the time of arrival, treatment with diuretics at the time of discharge, and history of CRI caused sizable changes in the parameter estimates for NIRxDM and IRxDM patients, suggesting that these factors accounted for much of the higher risk of heart failure.


View this table:
[in this window]
[in a new window]
  		Table 2 Risk for heart failure adjusted for prognostic factors

 
Readmission for Recurrent Myocardial Infarction
At 1 year after the myocardial infarction, 135 patients had been readmitted to the hospital because of recurrent myocardial infarction: 79 (6%) of the patients without diabetes, 34 (11%) of the patients with NIRxDM, and 22 (12%) of the patients with IRxDM (P = .001) The risk for readmission because of recurrent myocardial infarction for the 3 groups during the year after myocardial infarction is shown in Figure 2Go. Patients with NIRxDM and IRxDM had similar patterns of readmission (P < .001). Unadjusted estimates of relative risk were 1.84 (95% CI=1.23–2.74) for patients with NIRxDM and 2.02 (95% CI = 1.26–3.25) for patients with IRxDM. After age was controlled for, compared with the risk in patients without diabetes, the relative risk was 1.87 (95% CI = 1.25–2.80) for patients with NIRxDM and 2.25 (95% CI = 1.40–3.63) for patients with IRxDM.



View larger version (8K):
[in this window]
[in a new window]
  		Figure 2 Cumulative rates for readmission for recurrent myocardial infarction.

 
With adjustments for comorbid conditions related to diabetes mellitus, clinical findings on arrival at the hospital, and characteristics of the myocardial infarction in the multivariate model (Table 3Go), risk of recurrent myocardial infarction in patients with NIRxDM (RR = 1.52; 95% CI = 1.01–2.29) and IRxDM (RR = 1.40; 95% CI = 0.84–2.31) dramatically decreased. Increasing age, previous myocardial infarction and PVD, increased heart rate and comatose mental status at the time of arrival at the hospital, and treatment with diuretics and nitrates at the time of discharge were associated with an increased risk of recurrent myocardial infarction. In additional analyses, previous myocardial infarction and PVD, heart rate at the time of arrival, and treatment with diuretics and nitrates at the time of discharge caused large changes in parameter estimates, suggesting that these factors accounted for much of the elevated risk of recurrent myocardial infarction in patients with diabetes, particularly patients with IRxDM.


View this table:
[in this window]
[in a new window]
  		Table 3 Risk for recurrent myocardial infarction adjusted for prognostic factors

 
One-Year Mortality
At 1 year after the myocardial infarction, 308 patients had died: 198 (16%) of the patients without diabetes, 72 (24%) of the patients with NIRxDM, and 38 (21%) of the patients with IRxDM (P = .003). The risk for mortality among the 3 groups during the year after myocardial infarction is shown in Figure 3Go. Death was more likely in patients with NIRxDM than in patients with IRxDM or patients without diabetes mellitus (P = .003). Unadjusted relative risk estimates were 1.55 (95% CI = 1.18–2.03) for patients with NIRxDM and 1.35 (95% CI = 0.95–1.91) for patients with IRxDM. After age was controlled for, compared with the risk in patients without diabetes, the relative risk for death was 1.63 (95% CI = 1.24–2.13) in patients with NIRxDM and 1.74 (95% CI = 1.22–2.48) in patients with IRxDM. With adjustments for comorbid conditions related to diabetes mellitus, clinical findings on arrival at the hospital, and characteristics of the myocardial infarction in the multivariate model (Table 4Go), risk for death in patients with NIRxDM (RR= 1.08; 95% CI = 0.82–1.43) and IRxDM (RR = 0.92; 95% CI = 0.62–1.35) decreased dramatically. Increasing age; previous heart failure, stroke, CRI and PVD; the presence of heart failure, lower systolic blood pressure, higher heart rate, and elevated potassium levels at the time of arrival; lower LVEF and heart failure later during the hospitalization; myocardial infarction at an undetermined site; and treatment with diuretics at the time of discharge were associated with an increased risk of death in the following year. Treatment with ßblockers at the time of discharge was associated with a decreased likelihood of death. Inclusion of these prognostic factors eliminated the increased risk of patients with both NIRxDM and IRxDM. In further analyses, comorbid conditions related to diabetes mellitus (ie, previous heart failure, stroke, CRI, and PVD), heart failure and heart rate on arrival at the hospital, potassium levels within the first 24 hours of admission, LVEF, and treatment with diuretics at the time of discharge accounted for much of the increased risk.



View larger version (9K):
[in this window]
[in a new window]
  		Figure 3 Cumulative rates for mortality 1 year after myocardial infarction.

 

View this table:
[in this window]
[in a new window]
  		Table 4 Risk of mortality 1 year after myocardial infarction adjusted for prognostic factors

 

   Discussion
Top
 Abstract
 Methods
 Results
 Discussion
Clinical Implications
 Limitations
 Conclusions
 References
 
We found that compared with elderly patients without diabetes, elderly patients with NIRxDM and IRxDM have a significantly greater risk for readmission for heart failure and recurrent myocardial infarction, and that risk was greater in patients with IRxDM than in patients with NIRxDM. The elevated risk in patients with diabetes mellitus was partially explained by differences in comorbid conditions related to diabetes mellitus (previous CRI, myocardial infarction, and PVD); clinical findings on arrival at the hospital (heart failure and increased heart rate); and characteristics of the myocardial infarction, specifically measures of ventricular function (heart failure later during hospitalization, LVEF, and need for treatment with diuretics at the time of discharge). However, IRxDM remained an independent predictor of heart failure, and NIRxDM remained an independent predictor of recurrent myocardial infarction.

The basis for comparing our rates of heart failure and recurrent myocardial infarction with those found in other studies is limited, because these 2 outcomes have not been widely reported in elderly patients with diabetes mellitus, and differences between patients with NIRxDM and those with IRxDM have not been explored. Not unexpectedly, rates of heart failure in our study were lower than those reported in one series2 (30.1%) in which patients all had Q-wave myocardial infarctions. Rates were higher in our patients than in the patients in the Framingham Study1 (9%–16%), who were younger than our patients. Rates of recurrent myocardial infarction, although more widely studied, have also varied considerably, with some researchers3–5 finding higher (21–46%) rates and others1,7 finding lower rates (10%). Similar rates of reinfarction (18%) and heart failure (20%) occurred in patients with diabetes in the Diabetes Insulin-Glucose in Acute Myocardial Infarction study,35 and similarly elevated risk estimates have been reported for patients with unstable angina or non–Q-wave myocardial infarction.36

Although we found that elderly patients with diabetes mellitus have a significantly greater risk for mortality in the year after myocardial infarction than do patients without diabetes, mortality was greater in patients with NIRxDM than in those with IRxDM. In addition, the increased risk was explained by comorbid conditions related to diabetes mellitus and poorer ventricular function; neither NIRxDM nor IRxDM was an independent predictor of mortality 1 year after the myocardial infarction. Compared with the 6-month or 1-year mortality rates used in other studies, our mortality rates were higher in patients without diabetes (16% vs 3%–11%) and in patients with diabetes mellitus (21%–24% vs 2%–18%) than were estimates from clinical trials7,20,22 and were lower than mortality rates found in studies not based on clinical trials.4,23 Differences in mortality in previous studies may be due to the greater prevalence of Q-wave myocardial infarction, less aggressive management before the thrombolytic era,4,5,14,37 or the younger age of patients7,20,22 in those studies. Older age has consistently been associated with poorer outcomes both after myocardial infarction38,39 and in patients with diabetes mellitus1,2,4,7–12,16,17,19–22,25,36,40 and is associated with development of heart failure.17,38,39,41

In adjusted analysis, our mortality estimates were similar to those of some previous investigations10,16,20,24–26 and higher1,11,12,22 and lower9,19 than the risk estimates of others. This discrepancy may be explained by several factors. A variety of variables were included in previous multivariate models, a situation that would have affected the adjusted risk estimates. Individual risk estimates for factors included in previous models were calculated in only a few previous studies,16,19,25,36 and factors related to the acute myocardial infarction were reported in only a few investigations.16,25,36 In other studies, the researchers did not control for baseline differences between patients with and without diabetes mellitus5,13–15,18 or did not present adjusted risk estimates.2–4,8,21 Length of follow-up and the age of patients may also have influenced the risk estimates. Nevertheless, our finding of an elevated long-term mortality risk in patients with diabetes mellitus is consistent with the findings of previous studies.1,7,9–12,16,17,19,20,22,23,25,26,36

Data are limited on how the type of treatment of diabetes mellitus affects heart failure, recurrent myocardial infarction,1,6 and mortality.7,17,20–22 Unlike our results, those of several studies17,20,22,26 indicated that the unadjusted mortality risk was higher in patients with IRxDM than in patients with NIRxDM. When data were stratified by sex, others found that the increased mortality risk was confined to men with non–insulin-dependent diabetes mellitus21 or to women with insulin-dependent diabetes mellitus.7,21 We did not find any differences between men and women. These discrepancies between studies may be related to the younger age of the patients in previous studies; patients with diabetes mellitus tended to be older than patients without diabetes mellitus. In our study, patients with NIRxDM and IRxDM were younger than the patients without diabetes mellitus.

The age of the patients in our investigation and the high prevalence of diabetes mellitus (28%), compared with the prevalence found by others (5%–23%), suggest that our patients, whether classified as IRxDM or NIRxDM, probably had type 2 diabetes mellitus. In studies with younger patients, those patients classified as having insulin-dependent diabetes mellitus or IRxDM may have included more patients with type 1 diabetes mellitus. Sex-related differences are more important in younger patients both with42 and without43,44 diabetes and therefore may be more relevant to younger patients with type 1 diabetes mellitus. However, diabetes is an important cardiac risk factor in women, and the role of sex as a prognostic factor deserves further study.36

Our findings suggest that important differences exist between patients with NIRxDM and patients with IRxDM and that diabetes has an independent effect on heart failure and recurrent myocardial infarction but not on mortality. Although comparisons with previous studies according to classification of diabetes mellitus are not possible, it is important to determine how management of diabetes affects patients with type 2 diabetes mellitus. Multiple factors affect whether patients with type 2 diabetes mellitus are treated with insulin or not: physicians’ preferences, patients’ acceptance, the severity of the disease, and the occurrence of complications related to diabetes mellitus.

In the Diabetes Insulin-Glucose in Acute Myocardial Infarction study, the intervention group who maintained tighter control of blood glucose during the initial period of myocardial infarction and after discharge from the hospital had lower rates of heart failure (17% vs 22%) and mortality (10% vs 15%; P < .05) than did a control group who received usual care.35 Levels of blood glucose and hemoglobin A1c were important predictors of subsequent mortality.45 In addition, the greatest reductions in hemoglobin A1c levels and the most pronounced decrease in risk occurred in patients without prior insulin treatment and low levels of cardiac risk factors.45

We found that comorbid conditions related to diabetes mellitus (previous heart failure, myocardial infarction, CRI, PVD, and stroke) were not only important prognostic factors, but also contributed to the increased risk of heart failure (previous CRI), recurrent myocardial infarction (previous myocardial infarction and PVD), and death (previous heart failure, stroke, CRI, and PVD) observed in both patients with NIRxDM and patients with IRxDM. Although these comorbid conditions have been reported in only a few previous investigations of the outcomes of myocardial infarction,24,36 our findings are consistent with results that indicated the overall importance of heart failure, CRI, stroke, myocardial infarction, and PVD in patients with diabetes.29,46–51

Increased heart rate was not only an important prognostic factor but also attenuated the effect of NIRxDM and IRxDM on the outcomes we examined. Although the cause of the higher heart rate could not be determined in our study, other investigators26,42 found higher heart rates in similar patients at the time of arrival at the hospital, and patients with diabetes mellitus often have a decrease in heart rate variability because of autonomic dysfunction. Autonomic dysfunction is associated with metabolic control, sudden death, heart failure, and mortality.52–56 A similar percentage of patients with (22%) and without diabetes mellitus (21%) in our study were taking ß-blockers before the hospitalization, and inclusion of previous treatment in any of the multivariate models did not alter the influence of heart rate. Other factors, such as heart failure, however, may have accounted for the higher heart rate.

We found that heart failure at the time of arrival at the hospital, lower LVEF, and treatment with nitrates and diuretics at the time of discharge were not only important prognostic factors but also accounted for much of the increased risk in patients with NIRxDM and IRxDM. The occurrence of heart failure, despite similar LVEFs in patients with and without diabetes, may be related to diastolic dysfunction or to fluid overload associated with renal dysfunction. Older age is also related to the development of heart failure when systolic function is normal, and the prevalence of heart failure in elderly patients with myocardial infarction increases with increasing age.57,58 Other researchers have also found that heart failure on arrival or during the hospitalization is an important prognostic factor in patients with diabetes1,7,9,10,16,20–22,24,25,35,45,59,60 and that when heart failure was controlled for, the difference in mortality in patients with diabetes mellitus was attenuated.1,2,14

Whether potassium levels were related to previous use of angiotensin-converting enzyme inhibitors, digoxin, or diuretics; previous renal disease; or metabolic effects during the period of the acute myocardial infarction cannot be determined. However, potassium level was the only laboratory value associated with heart failure and mortality. Further investigation of the role of potassium is warranted. The need for use of diuretics and nitrates at the time of discharge probably reflected previous hypertension and heart failure, along with heart failure during the acute myocardial infarction. Others24 also found that treatment with diuretics and digitalis at the time of discharge was associated with an increased risk for mortality. These factors, however, are all indicators of ventricular function, and the findings not only indicate their importance as prognostic factors but also the large contribution of ventricular function to subsequent heart failure, recurrent myocardial infarction, and mortality in elderly patients with NIRxDM and IRxDM.


   Clinical Implications
Top
 Abstract
 Methods
 Results
 Discussion
 Clinical Implications
Limitations
 Conclusions
 References
 
Our findings have important implications for nurses practicing in a variety of settings. The prognostic factors can be used to target patients with NIRxDM and IRxDM who are at high risk for heart failure (increased age, previous coronary artery bypass or CRI, poorer mental status, heart failure or increased heart rate at the time of arrival at the hospital, lower LVEF or heart failure later during hospitalization, and use of diuretics at the time of discharge), recurrent myocardial infarction (increased age, previous myocardial infarction or PVD, increased heart rate or poorer mental status at the time of arrival, and use of diuretics or nitrates at the time of discharge), and mortality (increased age; previous heart failure, stroke, CRI, or PVD; lower systolic blood pressure; increased heart rate or higher potassium levels at the time of arrival; lower LVEF; undetermined site of myocardial infarction; heart failure later during hospitalization; and use of diuretics at the time of discharge) before the patients are discharged from the hospital after treatment for the acute myocardial infarction. Closer follow-up by visiting nurses or other care providers can then be instituted. These patients, along with their family members, may also benefit from additional education about signs and symptoms of heart failure and myocardial infarction before discharge. Referral to a cardiac rehabilitation program, with an emphasis on control of blood glucose levels, control of cardiac risk factors, and exercise capacity may lead to improved outcomes.61

Although heart failure at the time of arrival at the hospital and need for treatment with diuretics and nitrates at the time of discharge probably are the combined effects of preexisting ventricular dysfunction, as well as the effects of the acute myocardial infarction, these factors should alert healthcare providers to the possibility of a more complicated course after discharge. Additional study of the importance of heart rate, including heart rate variability, in elderly patients with diabetes is warranted. The prognostic and explanatory factors we describe here should serve as a basis for designing intervention studies aimed at lessening the substantial morbidity and mortality in this high-risk population.

Comorbid conditions related to diabetes mellitus (previous stroke, CRI, PVD, myocardial infarction, and heart failure) that attenuated the association between NIRxDM and IRxDM and these outcomes provide a basis for primary, secondary, and tertiary preventive measures. These comorbid conditions are all strongly associated with diabetes. Although stroke, PVD, heart failure, and myocardial infarction are not directly related to control of blood glucose levels, CRI has been linked to poor diabetes management. Recent evidence35,45 also supports the importance of glucose control during and after the acute myocardial infarction. In addition, diabetes is an important prognostic factor, even in patients without known heart disease29,36 and in patients with unstable angina.62 Management of diabetes and cardiac risk factors both before and after myocardial infarction therefore should be a priority for all healthcare providers.

The importance of lowering blood pressure63 and controlling lipid levels64,65 in patients with diabetes is a consistent finding. Despite our conservative estimate of the prevalence of diabetes mellitus and the outcomes we studied, both NIRxDM and IRxDM were associated with heart failure, recurrent myocardial infarction, and mortality. Although we could not differentiate type 1 from type 2 diabetes or assess the severity of diabetes or efficacy of diabetes treatment, either before or after the myocardial infarction, the long-term effects of glucose management on development of myocardial infarction and outcomes after myocardial infarction require further investigation.

Although the management of patients with myocardial infarction did not differ according to their diabetes status and the patients in our study were cared for several years ago, more current estimates continue to confirm the underuse of ß-blockers66–68 and of aspirin in patients with myocardial infarction or diabetes.69,70 Treatment with ß-blockers67,71 and angiotensin-converting enzyme inhibitors72,73 reduces mortality after myocardial infarction in patients with diabetes. Aspirin also reduces cardiac events in patients with diabetes74–76; however, in patients with coronary artery disease, aspirin may decrease the effectiveness of angiotensin-converting enzyme inhibitors.77 Recent studies also indicated the effectiveness of clopidogrel78 and tirofiban79 in reducing cardiac events, and treatment with antithrombotic agents is especially important in patients with diabetes.


   Limitations
Top
 Abstract
 Methods
 Results
 Discussion
 Clinical Implications
 Limitations
Conclusions
 References
 
One limitation of our study is that all data were obtained via a retrospective review of medical records and that hospitalization for heart failure and recurrent myocardial infarction was used as the basis for determining which patients experienced these 2 events. Silent myocardial infarctions might not have been detected, and patients might have died of heart failure or myocardial infarction before they reached the hospital. In addition, diagnoses of heart failure and recurrent myocardial infarction were not verified in patients readmitted because of these 2 abnormalities. Heart failure not requiring hospitalization was not detected, a situation that might have led to an underestimation of prognostic factors and of the outcomes we studied.

Diabetes mellitus was not confirmed by measurement of fasting plasma glucose levels. Exclusion of patients whose diabetes mellitus status could not be determined or who may have had diabetes mellitus diagnosed during the hospital stay, in combination with the higher threshold used for diagnosis before 1997,80 might have resulted in an underestimation of the prevalence of diabetes mellitus. The prevalence in our study, however, was greater than the prevalence in most previous investigations. However, some patients with diabetes mellitus might have been classified as not having diabetes mellitus, and we therefore may have underestimated the impact of diabetes mellitus. Other researchers have reported that many patients with coronary artery disease have undiagnosed diabetes. In addition, glucose abnormalities in patients with coronary artery disease are common, and an adverse effect of both diabetes and previously undiagnosed impairments in fasting blood glucose levels (fasting blood glucose, 6.1–7.0 mmol/L [110–126 mg/dL]) on long-term outcomes has been reported.81,82

We relied on LVEF determinations, clinical evidence of heart failure, and use of diuretics and nitrates at the time of discharge to assess ventricular function. Measurement of diastolic function would have provided important information on the patients in our study. Additional prognostic factors, such as silent myocardial infarction, heart failure after discharge, residual myocardial ischemia, LVEF, thrombogenic risk factors, and autonomic imbalance, that might have contributed to these outcomes83 were not assessed after the patients were discharged from the hospital.

In addition, because we used total CK levels rather than CK-MB levels, we may have overestimated infarct size, and criteria for diagnosis of myocardial infarction were fairly liberal. Although thrombolytic therapy was used less often in patients with diabetes, probably because of non–Q-wave myocardial infarction, and no data support the notion that thrombolytics are less effective in patients with diabetes,84 we did not have information on the patency of vessels.

Despite these limitations, this study has several important strengths. Unlike many previous studies, it was specifically designed to examine the care and outcomes of patients with diabetes mellitus. It was not an analysis of preexisting data. Our results provide current estimates of heart failure, recurrent myocardial infarction, and mortality in elderly patients with diabetes mellitus.

The availability of the records of the entire population of patients with myocardial infarction in Connecticut increases the generalizability of our findings. Finally, because detailed clinical information on the acute myocardial infarction period was available, we were able to examine the effect of NIRxDM and IRxDM on heart failure, recurrent myocardial infarction, and mortality.


   Conclusions
Top
 Abstract
 Methods
 Results
 Discussion
 Clinical Implications
 Limitations
 Conclusions
References
 
Similar to most recent studies on myocardial infarction in patients with diabetes mellitus,23,25,26,36 our results indicated an association between NIRxDM and IRxDM and increased risk of heart failure, recurrent myocardial infarction, and mortality. Comorbid conditions related to diabetes mellitus (previous myocardial infarction, heart failure, CRI, PVD, and stroke), clinical findings on arrival at the hospital (increased heart rate and occurrence of heart failure), and poorer ventricular function (decreased LVEF and use of diuretics at the time of discharge), accounted for much of the increased risk. The prognostic factors we described can be used before discharge from the hospital to target patients with NIRxDM and IRxDM who are at high risk for heart failure, recurrent myocardial infarction, and mortality. In addition, the contribution of comorbid conditions associated with diabetes mellitus to poorer long-term outcomes supports the need for aggressive control of blood glucose levels and cardiac risk factors, particularly dyslipidemias and blood pressure in patients with diabetes mellitus.


   ACKNOWLEDGMENTS
 
The analyses upon which this publication is based were performed under Contract No. 500-96-P549, "Utilization and Quality Control Peer Review Organization for the State of Connecticut," sponsored by the Health Care Financing Administration, Department of Health and Human Services. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organization imply endorsement by the US Government.

The authors assume full responsibility for the accuracy and completeness of the ideas presented. This article is a direct result of the Health Care Quality Improvement program initiated by the Health Care Financing Administration, which has encouraged identification of quality improvement projects derived from analysis of patterns of care, and therefore required no special funding on the part of the contractor. Ideas and contributions to the author concerning experience in engaging with issues presented are welcomed.

To purchase reprints, contact The InnoVision Group, 101 Columbia, Aliso Viejo, CA 92656. Phone, (800) 809-2273 or (949) 362-2050 (ext 532); fax, (949) 362-2049; e-mail, reprints{at}aacn.org.


   REFERENCES
Top
 Abstract
 Methods
 Results
 Discussion
 Clinical Implications
 Limitations
 Conclusions
 References
 

  1. Abbott RD, Donahue RP, Kannel WB, Wilson PWF. The impact of diabetes on survival following myocardial infarction. JAMA. 1988;260:3456–3460.[Abstract]
  2. Stone PH, Muller JE, Hartwell T, et al. The effect of diabetes mellitus on prognosis and serial left ventricular function after acute myocardial infarction: contribution of both coronary disease and diastolic left ventricular dysfunction to the adverse prognosis. The MILIS Study Group. J Am Coll Cardiol. 1989;14:49–57.[Abstract]
  3. Ulvenstam G, Åberg A, Bergstrand R, et al. Long-term prognosis after acute myocardial infarction in men with diabetes. Diabetes. 1985;34:787–792.[Abstract]
  4. Malmberg K, Rydén L. Myocardial infarction in patients with diabetes mellitus. Eur Heart J. 1988;9:256–264.
  5. Herlitz J, Malmberg K, Karlson BW, Ryden L, Hjalmarson A. Mortality and morbidity during a five-year follow-up of diabetics with myocardial infarction. Acta Med Scand. 1988;244:31–38.
  6. Milo C, Reicher-Reiss H, Goldbourt U, Boyko V, Behar S. Comparison of prognosis of acute myocardial infarction in insulin-treated diabetic women versus men. The SPRINT Study Group. Am J Cardiol. 1994;74:1275–1276.[Medline]
  7. Behar S, Boyko V, Reicher-Reiss H, Goldbourt U. Ten-year survival after acute myocardial infarction: comparison of patients with and without diabetes. Am Heart J. 1997;133:290–296.[Medline]
  8. Singer DE, Moulton AW, Nathan DM. Diabetic myocardial infarction: interaction of diabetes with other preinfarction risk factors. Diabetes. 1989;38:350–357.[Abstract]
  9. Sprafka JM, Burke GL, Folsom AR, McGovern PG, Hahn LP. Trends in prevalence of diabetes mellitus patients with myocardial infarction and effect of diabetes on survival: The Minnesota Heart Survey. Diabetes Care. 1991;14:537–543.[Abstract]
  10. Donahue RP, Goldberg RJ, Chen Z, Gore JM, Alpert JS. The influence of sex and diabetes mellitus on survival following acute myocardial infarction: a community-wide perspective. J Clin Epidemiol. 1993;46:245–252.[Medline]
  11. Orlander PR, Goff DC, Morrissey M, et al. The relation of diabetes to the severity of acute myocardial infarction and post-myocardial infarction survival in Mexican-Americans and non-Hispanic whites: The Corpus Christi Heart Project. Diabetes. 1994;43:897–902.[Abstract]
  12. Abbud ZA, Shindler DM, Wilson AC, Kostis JB. Effect of diabetes mellitus on short- and long-term mortality rates of patients with acute myocardial infarction: a statewide study. Am Heart J. 1995;130:51–58.[Medline]
  13. Soler NG, Bennett MA, Pentecost BL, Fitzgerald MG, Malins JM. Myocardial infarction in diabetics. Q J Med. 1975;54:125–132.
  14. Jaffe AS, Spadaro JJ, Schechtman K, Roberts R, Geltman EM, Sobel BE. Increased congestive heart failure after myocardial infarction of modest extent in patients with diabetes mellitus. Am Heart J. 1984;108:31–37.[Medline]
  15. Weitzman S, Wagner GS, Heiss G, Haney TL, Slome C. Myocardial infarction site and mortality in diabetes. Diabetes Care. 1982;5:31–35.[Abstract]
  16. Martin CA, Thompson PL, Armstrong BK, Hobbs MST, de Klerk N. Long-term prognosis after recovery from myocardial infarction: a nine year follow-up of the Perth Coronary Register. Circulation. 1983;68:961–969.[Abstract/Free Full Text]
  17. Smith SC, Gilpin E, Ahnve S, et al. Outlook after acute myocardial infarction in the very elderly compared with that in patients aged 65 to 75 years. J Am Coll Cardiol. 1990;16:784–792.[Abstract]
  18. Molstad P, Nustad M. Acute myocardial infarction in diabetic patients. Acta Med Scand. 1987;222:433–437.[Medline]
  19. Wong ND, Cupples LA, Ostfeld AM, Levy D, Kannel WB. Risk factors for long-term coronary prognosis after initial myocardial infarction: the Framingham Study. Am J Epidemiol. 1989;130:469–480.[Abstract/Free Full Text]
  20. Barbash GI, White HD, Modan M, Van de Werf F. Significance of diabetes mellitus in patients with acute myocardial infarction receiving thrombolytic therapy. J Am Coll Cardiol. 1993;22:707–713.[Abstract]
  21. Zuanetti G, Latini R, Maggioni AP, Santoro L, Franzosi MG. Influence of diabetes on mortality in acute myocardial infarction: data from the GISSI-2 Study. J Am Coll Cardiol. 1993;22:1788–1794.[Abstract]
  22. Mak KH, Moliterno DJ, Granger CB, et al. Influence of diabetes mellitus on clinical outcome in the thrombolytic era of acute myocardial infarction. J Am Coll Cardiol. 1997;30:171–179.[Abstract]
  23. Miettinen H, Lehto S, Salomaa V, et al. Impact of diabetes on mortality after the first myocardial infarction. The FINMONICA Myocardial Infarction Register Study Group. Diabetes Care. 1998;21:69–75.[Abstract]
  24. Herlitz J, Karlson BW, Karlsson T, Lindqvist J, Sjolin M. Predictors of death during 5 years after discharge among patients with a suspected acute coronary syndrome with particular emphasis on whether an infarction was developed. Int J Cardiol. 1998;66:73–80.[Medline]
  25. Melchior T, Kober L, Madsen CR, et al. Accelerating impact of diabetes mellitus on mortality in the years following an acute myocardial infarction. TRACE Study Group. Trandolapril Cardiac Evaluation. Eur J Cardiol. 1999;20:973–978.
  26. Berger AK, Breall JA, Gersh BJ, et al. Effects of diabetes mellitus and insulin use on survival after acute myocardial infarction in the elderly (the Cooperative Cardiovascular Project). Am J Cardiol. 2001;87:272–277.[Medline]
  27. Harris MI, Flegal KM, Cowie CC, et al. Prevalence of diabetes, impaired fasting glucose tolerance in US adults: the Third National Health and Nutrition Examination Survey, 1988–1994. Diabetes Care. 1998;21:518–524.[Abstract]
  28. 2000 Heart and Stroke Facts and Statistics. Dallas, Tex: American Heart Association; 2000.
  29. Haffner SM, Lehto S, Ronnemaa T, Pyorala K, Laakso M. Mortality from coronary heart disease in subjects with type 2 diabetes and in nondiabetic subjects with and without prior myocardial infarction. N Engl J Med. 1998;339:229–234.[Abstract/Free Full Text]
  30. Ellerbeck EF, Jenks SF, Radford MJ, et al. Quality of care for Medicare patients with acute myocardial infarction: a four-state pilot study from the Cooperative Cardiovascular Project. JAMA. 1995;273:1509–1514.[Abstract]
  31. Mantel N. Chi-square tests with one degree of freedom: extensions of the Mantel-Haenszel procedure. J Natl Cancer Inst. 1963;58:690–700.
  32. Kaplan EL, Meier P. Nonparametric estimation from incomplete observation. J Am Stat Assoc. 1958;53:457–481.
  33. Cox DR. Regression models and life tables (with discussion). J R Stat Soc. 1972;34:187–220.
  34. Rothman KJ. Analysis with multiple levels of exposure. In: Modern Epidemiology. Boston, Mass: Little Brown & Co Inc; 1986.
  35. Malmberg K, Ryden L, Hamsten A, Herlitz J, Waldenstrom A, Wedel H. Effects of insulin treatment on cause-specific one-year mortality and morbidity in diabetic patients with acute myocardial infarction. DIGAMI Study Group. Diabetes Insulin-Glucose in Acute Myocardial Infarction [see comments]. Eur Heart J. 1996;17:1337–1344.[Abstract/Free Full Text]
  36. Malmberg K, Yusuf S, Gerstein HC, et al. Impact of diabetes on long-term prognosis in patients with unstable angina and non-Q-wave myocardial infarction: results of the OASIS (Organization to Assess Strategies for Ischemic Syndromes) Registry. Circulation. 2000;102:1014–1019.[Abstract/Free Full Text]
  37. Soler NG, Pentecost BL, Bennett MA, FitzGerald MG, Lamb P, Malins JM. Coronary care for myocardial infarction in diabetics. Lancet. 1974;1:475–477.[Medline]
  38. Goldberg RJ, Gore JM, Gurwitz JH, et al. The impact of age on the incidence and prognosis of initial acute myocardial infarction: the Worcester Heart Attack Study. Am Heart J. 1989;117:543–549.[Medline]
  39. Rich MW, Bosner MS, Chung MK, Shen J, Makenzie JP. Is age an independent predictor of early and late mortality in patients with acute myocardial infarction? Am J Med. 1992;92:7–13.[Medline]
  40. Malmberg K, McGuire DK. Diabetes and acute myocardial infarction: the role of insulin therapy. Am Heart J. 1999;138:S381–S386.[Medline]
  41. Marcus FI, Friday K, McCans J, et al. Age-related prognosis after acute myocardial infarction (the Multicenter Diltiazem Postinfarction Trial). Am J Cardiol. 1990;65:559–566.[Medline]
  42. Vaccarino V, Parsons L, Every NR, Barron HV, Krumholz HM. Impact of history of diabetes mellitus on hospital mortality in men and women with first acute myocardial infarction. The National Registry of Myocardial Infarction 2 Participants. Am J Cardiol. 2000;85:1486–1489.[Medline]
  43. Vacarrino V, Parsons L, Every NR, Barron HV, Krumholz HM. Sex-based differences in early mortality after myocardial infarction. The National Registry of Myocardial Infarction 2 Participants. N Engl J Med. 1999;341:217–225.[Abstract/Free Full Text]
  44. Vacarrino V, Horwitz RI, Meehan TP, Petrillo MK, Radford MJ, Krumholz HM. Sex differences in mortality after acute myocardial infarction: evidence for a sex-age interaction. Arch Intern Med. 1998;158:2054–2062.[Abstract/Free Full Text]
  45. Malmberg K, Norhammar A, Wedel H, Ryden L. Glycometabolic state at admission: important risk marker of mortality in conventionally treated patients with diabetes mellitus and acute myocardial infarction: long-term results from the Diabetes Insulin-Glucose in Acute Myocardial Infarction (DIGAMI) study. Circulation. 1999;99:2626–2632.[Abstract/Free Full Text]
  46. Kannel WB, Hjortland M, Castelli WP. Role of diabetes in congestive heart failure: the Framingham Study. Am J Cardiol. 1974;34:29–34.[Medline]
  47. Sutton-Tyrrell K, Rihal C, Sellers MA, et al. Long-term prognostic value of clinically evident noncoronary vascular disease in patients undergoing coronary revascularization in the Bypass Angioplasty Revascularization Investigation (BARI). Am J Cardiol. 1998;81:375–381.[Medline]
  48. Schmitz A, Vaeth M. Microalbuminuria: a major risk factor in non-insulin-dependent diabetes. A 10-year follow-up study of 503 patients. Diabetic Med. 1988;5:126–134.[Medline]
  49. Mattock MB, Keen H, Viberti GC, et al. Coronary heart disease and urinary albumin excretion rate in type 2 (non-insulin-dependent) diabetic patients. Diabetologia. 1988;31:82–87.[Medline]
  50. Pyorala K, Pedersen TR, Kjekshus J, Faergeman O, Olsson AG, Thorgeirsson G. Cholesterol lowering with simvastatin improves prognosis of diabetic patients with coronary heart disease: a subgroup analysis of the Scandinavian Simvastatin Survival Study (4S). Diabetes Care. 1997;20:614–620.[Abstract]
  51. Iseki K, Fukiyama K. Long-term prognosis and incidence of acute myocardial infarction in patients on chronic hemodialysis. Am J Kidney Dis. 2000;36:820–825.[Medline]
  52. Bigger JT, Fleiss JL, Steinman RC, Rolnitzky LM, Kleiger RE, Rottman JN. Correlations among time and frequency domain measures of heart period variability two weeks after acute myocardial infarction. Am J Cardiol. 1992;69:891–898.[Medline]
  53. Kleiger RE, Miller JP, Bigger JT, Moss AJ. Decreased heart rate variability and its association with increased mortality after acute myocardial infarction. Am J Cardiol. 1987;59:256–262.[Medline]
  54. O’Brien IA, McFadden JP, Corrall RJM. The influence of autonomic neuropathy on mortality in insulin-dependent diabetes. Q J Med. 1991;79:495–502.[Medline]
  55. Ewing DJ, Campbell IW, Clarke BF. Mortality in diabetic autonomic neuropathy. Lancet. 1976;1:601–603.[Medline]
  56. Burger AJ, Hamer AW, Weinrauch LA, D’Elia JA. Relation of heart rate variability and serum lipoproteins in type 2 diabetes mellitus and chronic stable angina pectoris. Am J Cardiol. 1998;81:945–949.[Medline]
  57. Mayer-Oakes SA, Oye RK, Leake B. Predictors of mortality in older patients with acute pulmonary edema and normal ejection fraction after acute myocardial infarction. J Am Geriatr Soc. 1991;39:862–868.[Medline]
  58. Pernenkil R, Vinson JM, Shah AS, Beckham V, Wittenberg C, Rich MW. Course and prognosis in patients > or = 70 years of age with congestive heart failure and normal versus abnormal left ventricular ejection fraction. Am J Cardiol. 1997;79:216–219.[Medline]
  59. Smith JW, Marcus FI, Serokman R. Prognosis of patients with diabetes mellitus after acute myocardial infarction. Am J Cardiol. 1984;54:718–721.[Medline]
  60. Savage MP, Krolewski AS, Kenien GG, Lebeis MP, Christlieb AR, Lewis SM. Acute myocardial infarction in diabetes mellitus and significance of congestive heart failure as a prognostic factor. Am J Cardiol. 1988;62:665–669.[Medline]
  61. Yu CM, Lau CP, Cheung BMY, et al. Clinical predictors of morbidity and mortality in patients with myocardial infarction or revascularization who underwent cardiac rehabilitation, and importance of diabetes mellitus and exercise capacity. Am J Cardiol. 2000;85:344–349.[Medline]
  62. Fava S, Azzopardi J, Agius-Muscat H. Outcome of unstable angina in patients with diabetes mellitus. Diabet Med. 1997;14:209–213.[Medline]
  63. Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes: UKPDS 38. UK Prospective Diabetes Study Group. BMJ. 1998;317:703–713.[Abstract/Free Full Text]
  64. Randomized trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet. 1994;344:1383–1389.[Medline]
  65. Goldberg RB, Mellies MJ, Sacks FM, et al. Cardiovascular events and their reduction with pravastatin in diabetic and glucose-intolerant myocardial infarction survivors with average cholesterol levels: subgroup analyses in the cholesterol and recurrent events (CARE) trial. The Care Investigators. Circulation. 1998;98:2513–2519.[Abstract/Free Full Text]
  66. Krumholz HM, Radford MJ, Wang Y, Chen J, Heiat A, Marciniak T. National use and effectiveness of ß-blockers for the treatment of elderly patients with acute myocardial infarction: National Cooperative Cardiovascular Project. JAMA. 1998;280:623–629.[Abstract/Free Full Text]
  67. Gottlieb SS, McCarter RJ, Vogel RA. Effect of beta-blockade on mortality among high-risk and low-risk patients after myocardial infarction. N Engl J Med. 1998;339:489–497.[Abstract/Free Full Text]
  68. Youn