acute myocardial infarction in diabetes mellitus and significance of congestive heart failure as a...

Acute Myocardial Infarction in Diabetes Mellitus and Significance of Congestive Heart Failure as a

Prognostic Factor Michael P. Savage, MD, Andrzej S. Krolewski, MD, PhD, Gregory G. Kenien, MD,

Mark P. Lebeis, MD, A. Richard Christlieb, MD, and Stanley M. Lewis, MD

Diabetes mellitus has been associated with high mortality rates in patients with acute myocardial infarction (AMI). To better define prognosis in this population, the clinical course of 183 diabetics with AMI was studied. In-hospital mortality for all patients was 26% (52 of 183 patients). Mortality was significantly higher in patients with prior AMI than in patients without prior AMI (41 vs lS%, p <O.Ol) and was significantly higher in women than in men (37 vs lS%, p <O.Ol). The P-fold increase in mortality among diabetic women was observed both in patients with and without prior AMI. The excess mortality among diabetic women was attributable to their increased risk for severe congestive heart failure (CHF) and cardiogenic shock. Death due to CHF occurred in 22% of all diabetic women with AMI compared with 6% of the diabetic men (p <O.Ol). Death resulting from complications other than CHF was similar for both sexes. There were no male-female differences in the history of prior AMI, systemic hypertension, obesity, nephropathy, frequency of Q-wave AMI, anterior AMI or peak creatine kinase levels to account for the high risk for CHF in diabetic women. It is therefore conclud- ed that diabetic women with AMI are at increased risk for death due to CHF, and that this risk is not readily attributable to known conditions associated with CHF.

(Am J Cardiol 1988;62:665-669)

From the New England Deaconess Hospital, Joslin Diabetes Center, and Brigham and Women’s Hospital, Boston, Massachusetts, and the Thomas Jefferson University Hospital, Philadelphia, Pennsylvania. Manuscript received December 22, 1987; revised manuscript received and accepted May 31,1988.

Address for reprints: Michael P. Savage, MD, Cardiac Catheter- ization Laboratory, Suite 561 I-D, N.H., Thomas Jefferson University Hospital, Philadelphia, Pennsylvania 19107.

C oronary artery disease continues to be the princi- pal cause of death in patients with diabetes mellitus (DM).’ The increased cardiovascular risk

among diabetics results both from an increased incidence of disease due to accelerated atherogenesis and from greater case-fatality rates during clinical events. In the setting of acute myocardial infarction (AMI), DM has been associated with high mortality rates.2-6 How- ever, the causes of the increased mortality in diabetics with AM1 remain unknown and specific diabetic sub- groups at high risk remain to be identified. To better define prognosis in this population, we studied the clinical course of 183 diabetic patients with AMI.

METHODS The study population was comprised of 183 diabetic

patients diagnosed with AM1 at the New England Dea- coness Hospital between 1974 and 1979. Most were followed as outpatients at the Joslin Clinic for diabetic care before hospitalization. Data were obtained by re- view of all case records during the study period with diagnoses of DM and suspected AMI. To be included in the study, the diagnosis of DM had to be established at least 1 month before AM1 and AM1 had to be docu- mented by the presence of at least 2 of the following: prolonged chest pain, elevation in cardiac enzymes and typical electrocardiographic changes. All patients sur- viving more than 24 hours in-hospital had elevation in cardiac enzymes or postmortem evidence confirming the acute nature of the infarct. Patients transferred to the New England Deaconess Hospital after initial diagnosis of AM1 at another hospital were excluded from the study.

AMIs were classified as either Q-wave or non-Q- wave infarcts by the presence or absence, respectively, of Q-waves of 10.04 second in duration in at least 2 electrocardiographic leads. Complications of AM1 analyzed included congestive heart failure (CHF), ventricular tachycardia, ventricular fibrillation, supraventricular tachyarrhythmia, bradyarrhythmia and heart block. CHF was classified according to the criteria of Killip and Kimball.’ Ventricular tachycardia was defined as 13 consecutive beats of ventricular origin at a rate > 120 beats/mm8 Only primary ventricular tachycardia and ventricular fibrillation were evaluated for the pur- poses of this report-ventricular arrhythmias occurring only during terminal CHF or cardiogenic shock were considered part of the terminal event and were excluded from further analysis. Mortality was defined as death

THE AMERICAN JOURNAL OF CARDIOLOGY OCTOBER 1, 1988 666

MYOCARDIAL INFARCTION IN DIABETES

TABLE I Clinical Variables of Diabetic Patients with Acute Myocardial Infarction

Men Women All Patients (n = 88) (n = 95) (n = 183)

Medical history preceding hospitalization DM 110 years (%) 57 (65) 80 (84)’ 137 (75) Insulin therapy (%) 51(57) 76 (80)” 127 (69) Retinopathyt (%) ll(15) 19 (26) 30 (20) Nephropathyt (%) 31 (35) 26 (29) 57 (32) Obesity5 (%) 25 (32) 37 (44 62 (38) Systemic hypertension’1 (%) 48 (55) 40 w 88 (49) Angina pectoris (%) 40 VW 58 (63) 98 (56) Prior AMI (%) 41 (47) 42 (4.9 83 (45) Prior CHF (%) 26 (31) 48 (52)’ 74 (42)

Clinical features during acute myocardial infarction Chest pain (%) 69 (80) 68 (74) 137 (77) Diabetic ketoacidosis (%) 3 (3) 4 (4) 7 (4) Infarct type

Q-wave (%) 52 (59) e-4 (46) 96 (52) Non-Q-wave (%) 32 (36) 41 (43) 73 w Unclassified (%) 4 (5) 10 (11) 14 (8)

Anterior AMIn (%) 27 (52) 24 (55) 51(53)

* p <O.Ol in comparison with men; + includes patients with retinitis proliferans detected during or before hospitalizabon: f serum creatinine 21.5 mg/dl: *weight 2120% of ideal body weight: ” history of drug therapy for hypertension or blood pressure >160/95 or both on 3 or more occasions; ’ based on Q-wave AMI only.

AMI = acute myocardial infarction; CHF = congestive heart failure; DM = diabetes mellitus.

TABLE II Complications of Acute Myocardial Infarction

Men Women All Patients (n = 88) (n = 95) (n = 183)

Ventricular tachycardia or 13 (15) 14 (15) 27 (15) fibrillation (%)

Supraventricular 19 (21) 26 (27) 45 (24) tachyarrhythmias (%)

Second- or third-degree heart 12 (14) 21 (22) 33 (18) block (%)

Congestive heart failure (Killip class)

I (%) 31 (35) 19 (20) 50 (27) II (%) 35 m 34 (36) a 69(38)

Ill (%) 15 (17) 18 (19) 33 (18) IV (%) 7 (8) 24 (25) I 31 (17)

* p <O.Ol in comparison with men.

occurring during hospitalization (mean duration 18 days).

Statistical analysis: In the analysis of data, dichoto- mous variables were assessed by the chi-square method. The joint effects of potential risk factors on CHF and mortality were analyzed by multiple logistic regression.9

RESULTS The patient population consisted of 88 men and 95

women. Mean age was 66 years (range 30 to 88). Perti- nent features of prior medical history are listed in Table I. Mean duration of DM was 17 f 11 years. Diabetic therapy included insulin in 126 patients (69%), oral agents in 43 (23%) and diet alone in 14 (8%). Although there were no differences between the sexes in terms of systemic hypertension, obesity or previous AMI, diabetic women were more likely than men to have a past history of CHF. I

0 lIlll!-

lnf%s Initial

Infarcts RE:z?

The presence or absence of chest pain, diabetic FIGURE 1. Hospital mortality of diabetic patients with acute ketoacidosis as well as the type and location of AM1 are myocardial infarction.

666 THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 62

TABLE Ill Cause of Death

Men (n = 88)

Women (n = 95)

All Patients (n = 183)

Congestive heart failure (%) Bradyarrhythmia (%) Ventricular arrhythmia (%) Acute VSD or MR (%) Pulmonary embolism (%) Other (%)

5 (6) 21(22)U 26 (14) 5 (6) 4 (4) 9 (5) 5 (6) 2 (2) 7 (4) 0 (0) 2 (2) 2 (1) 0 (0) 2 (2) 2 (1) 2 (2) 4 (4) 6 (3)

* p <O.Ol in comparison with men. MR = mitral regurgitation; VSD = ventricular septal defect.

listed in Table I. Chest or referred pain was present in 77%. Diabetic ketoacidosis was infrequent (4%) despite the high prevalence of insulin-dependent DM. Q-wave AM1 was present in 52% of all patients and involved the anterior wall in 53%. There were no significant differences in these features between men and women.

The frequencies of ventricular arrhythmias, supraventricular arrhythmias and heart block are listed in Table II. There were no significant differences between men and women in the incidence of these arrhythmias. The distribution of CHF severity in the group as a whole and by gender is also listed in Table II. Class III or IV CHF was present in 44% of the diabetic women compared with only 25% of the diabetic men. Whereas there were no differences between the male and female groups in arrhythmic complications, the development of severe CHF was significantly greater in diabetic women.

Figure 1 shows the hospital mortality of diabetic men and women for all AMI, initial AM1 and recurrent AMI. Overall hospital mortality was 28% (52 of 183 patients). Mortality was significantly higher in patients with prior AM1 and in women. Mortality among patients suffering a recurrent AM1 was 41% in contrast to 18% in patients with their first AM1 (p <O.Ol). Mortal- ity among all diabetic women was 37% while among diabetic men it was 19% (p <O.Ol). This 2-fold increase

p<o.o, I I

I I I I

in mortality among diabetic women was observed both with initial and with recurrent AMI.

To better assess the increased risk in diabetic women, mortality was analyzed according to the cause of death. As listed in Table III, the excess mortality among diabetic women was attributable entirely to CHF. Death from CHF occurred in 22% of all women compared with 6% of all men (p <O.Ol), whereas overall mortality from causes other than CHF was identical for both sexes.

Figure 2 illustrates the important relation between CHF severity and prognosis. Prognosis was excellent in the absence of CHF (4% in-hospital mortality in Killip class I), was intermediate with moderate degrees of CHF (17 and 2 1% mortality in Killip classes II and III, respectively), and was almost uniformly fatal in the setting of cardiogenic shock (96% mortality in Killip class IV). Figure 2 also shows that, for any given degree of CHF, prognosis was the same for both sexes. Therefore, the increased risk of death among diabetic women appears to be related to their greater risk of developing severe CHF and not to additional factors independent of ventricular function.

Diabetic women had an increased risk of CHF and death during AMI, and were also more likely than their male counterparts to have prior histories of CHF, insulin therapy and DM durations of 110 years. Because of these differences in clinical background, multivariate analyses were performed to assess whether gender rep- resents an important independent risk factor for the development of CHF and death. To eliminate the effect of previous CHF, the risk of CHF during AM1 was analyzed only among patients without a prior history of CHF. Table IV lists the results of multiple logistic regression analysis on the risk of developing CHF in the absence of a previous history of CHF. Five independent variables were identified as significant determinants of new CHF: female sex, insulin therapy, age, history of systemic hypertension and Q-wave AMI. Because patients with prior CHF were specifically excluded, this analysis demonstrates that the excessive risk of developing CHF in diabetic women persists even after past history of CHF and other risk factors are accounted for.

Multivariate analysis was also used to assess the joint effects of different variables on the risk of mortality in all 183 patients (Table V). Four variables were significant determinants of mortality: female sex, insulin therapy, prior AM1 and Q-wave AMI. Thus, diabetic women carry an increased risk of CHF and death which is independent of other risk factors.

DISCUSSION Partamian and Bradley,2 in a study from our institu-

tion before the introduction of the coronary care unit, previously reported the mortality of diabetic patients with AMI. Comparison of the present study with this prior investigation, conducted 20 years earlier, suggests an improvement in prognosis over this time period. Mortality among the 258 patients in their study was 41% in contrast to the 28% mortality currently reported. Mortality among patients with initial AM1 decreased

TABLE IV Multiple Logistic Regression Analysis of the Joint Effects of Selected Variables Predictive of Congestive Heart Failure in Diabetics With Acute Myocardial Infarction Without Prior History of Heart Failure

Independent Variables* Relative Risk p Valuse -

Sex (men vs women) 2.88 <0.05 Diabetic therapy (huh vs other) 3.76 <O.Ol Age(70vs 55yrs) 2.76 <O.Ol Systemic hypertension Cyesvs no) 2.74 <0.05 Q-wave AMI (present vs absent) 6.70 -co.01

*Additional variables which did notcontribute to the regression model include: (1) prior AMI: (2) duration of diabetes: (3) obesity: (4) smoking; (5) nephropathy; (6) peak creatine kinasa level.

AMI = acute myocardial infarction. I

TABLE V Multiple Logistic Regression Analysis of the Joint Effects of Selected Variables Predictive of Hospital Mortality Among Diabetics With Acute Myocardial Infarction 1 Independent Variables* Relative Risk p Value 1

Sex (men vs women) Diabetic therapy (insulin vs other) Prior AMI (yesvs no) Q-wave AMI (present vs absent)

3.47 <O.Ol 2.60 <0.05 3.52 -co.01 4.85 <O.Ol

* Additional variables which did notsignificantly contribute to the regression model include: (1) prior congestive heart failure: (2) age; (3) duration of diabetes; (4) ;{zrtenslon; (5) smoking: (6) obesity; (7) nephropathy: (8) peak craatlne klnase .-.-.. 1

AMI = acute myocardial infarctron. I

from 38 to lB%, while mortality among patients with recurrent AM1 improved to a lesser degree from 55 to 41%. The present study analyzed the outcome of diabetics with AM1 prior to the utilization of thrombolysis and coronary angioplasty. Therefore, the effect of these acute interventions on the course of AM1 in diabetes is not assessed by this data.

Several investigators have shown that diabetic patients with AM1 have higher mortality rates than their nondiabetic counterparts.3-5 The reasons for this excess mortality are unknown, although Jaffe et allo noted

FIGURE 2. Hospital mortalii of diabetic patients wtth ecute myocardial infarction according to Killip class.


MYOCARDIAL INFARCTION IN DIABETES

an increased incidence of CHF among diabetics after AMI. Hospital mortality in the present study is compa- rable with the cumulative 31% mortality of diabetics in other studies from the coronary care unit era.3-5,31-14 Nevertheless, the patients in our study differ in important clinical respects from those of previous reports because there was a relatively high proportion of insulin- dependent and longstanding DM. Therefore, differences reflecting this patient population, most of whom were followed by the Joslin Clinic, must be considered when comparing present findings with other published data.

Four risk factors were identified by multivariate analysis as independent determinants of prognosis. In order of relative importance, they were as follows: presence of Q-wave AMI, history of prior AMI, female gender and insulin therapy before hospitalization. The higher mortality in diabetic patients with Q-wave AM1 and with recurrent AM1 concurs with previous studies of AM1 in the general population, and reflects the association of these factors with more extensive myocardial damage. l 5-20

Patients receiving insulin therapy were at increased risk for both CHF and death. The reason for the excess risk is unclear, but it may be due to more extensive coronary artery disease or to the more severe metabolic derangement in this group of diabetics. Because ischemic myocardium is dependent upon the anaerobic me- tabolism of glucose and because insulin promotes utilization of glucose, insulin deficiency could be particular- ly detrimental in the setting of AMI.21,22 The poor prognosis associated with insulin-dependent DM suggests a need for further investigation into the role of more intensive metabolic interventions in periinfarction management.

The most striking finding of this study was the 2- fold increase in mortality among diabetic women. A similar excess in mortality among diabetic women with AM1 was reported in a smaller study by Tansey et a1.13 The Framingham Study has also implicated diabetic women as a high-risk group: compared with diabetic men, diabetic women had a 2-fold increase in case fatality rates associated with “coronary events.“6 Such differences in prognosis have not been uniformly found. Czyzyk et al5 noted no difference in the clinical course of 93 male and 61 female diabetics with AM1 in War- saw.

The excess mortality among diabetic women with AM1 is due to their increased risk of developing CHF. The reasons why diabetic women develop more severe CHF remain unknown. One possibility is that diabetic women may have more extensive coronary artery disease which predisposes them to greater dysfunction in areas of noninfarcted myocardium or to larger infarct size due to compromised collateral circulation. Al- though there were no male-female differences in frequency of Q-wave AMI, anterior AM1 or peak creatine kinase levels to suggest larger infarct size in diabetic women, more sensitive estimates of infarct size are nec- essary to exclude this possibility. An alternative possibility could involve a diabetes-specific cardiomyopathy, which is characterized by pathologic changes in the

668 THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 62

microvasculature and myocardial interstitium and is un- related to large-vessel atherosclerosis.23 If diabetic women are prone to underlying cardiomyopathic disease, a higher incidence of overt CHF complicating AM1 would be expected. Adequate data to evaluate these hypotheses are not available.

The present study adds to a growing body of evidence indicating a selectively pernicious relation between diabetes and heart disease in women. As shown by the Framingham Study, the relative impact of DM on the development of coronary artery disease is sub- stantially greater in women than in men. After adjust- ing for other major risk factors, diabetic men carry a 70% excess risk of coronary heart disease while diabetic women carry a 200% excess risk.6 Whereas cardiovascular disease predominantly affects males in the nondiabetic population, DM negates the relative immunity from atherosclerosis bestowed upon the premenopausal female.1,24 The etiology of this sexual discrimination in cardiovascular risk appears to be mediated in part by the unfavorable lipoprotein profile of diabetic women.6,25

The Framingham Study also examined the role of gender on the risk of myocardial failure in DM.26 While diabetic men had a 2-fold increase in frequency of CHF compared to their nondiabetic counterparts, diabetic women suffered a 5-fold increased risk. Moreover, the increased risk of CHF in diabetics occurred both in the presence and absence of coronary artery disease. Among diabetics with coronary artery disease, women had an incidence of CHF over twice that of men. These findings help account for the high incidence of prior CHF observed in the diabetic women in our study. The Framingham Study thus indicated an untoward interac- tion between DM and gender, whereby diabetic women carry an excessive risk of CHF. The findings of the present study, confined to the setting of AMI, complement this earlier report. The risk of CHF among diabetic women is substantial. Further investigation into the cause of this lethal association is needed.

Acknowledgment: We wish to thank Drs. Robert F. Bradley, 0. Stevens Leland and James H. Warram for help in the formative stage of this project, Cathy Karp for assistance with data processing, and Cathy Franzeo and Laraine Bartlett for excellent secretarial assistance.

REFERENCES 1. Kmlewski AS, Warram JH, Christlieb AR. Onset, course, complications and prognosis of diabetes mellitus. In: Marble A, Krall L, Bradley RF, Christlieb AR, Soeldner JS, eds. Joslin’s Diabetes Mellitus. 12th ed. Philadelphia: Lea & Fe- biger, 1985:251-277. 2. Partamian JO, Bradley RF. Acute myocardial infarction in 258 cases of diabetes: immediate mortality and five-year survival. PI Engl J Med 1965; 273:455-461. 3. Soler NG, Bennett MA, Pentecost BL, Fitzgerald MG, Malins JM. Myocardi- al infarction in diabetics. Q J Med 1975;44:125-132. 4. Harrower AD, Clarke BF. Experience of coronary care in diabetes. Br Med J 1974;1:126-128. 5. Czyzyk A, Krolcwski AS, Szablowska S, Alot A, Kopczynski J. Clinical course of myocardial infarction among diabetic patients. Diabetes Care 2980;3:526-529. 6. Kannel WB. Lipids, diabetes and coronary heart disease: insights from the Framingham Study. Am Heart J 1985:110:1100-1107. 7. Killip T, Kimball JT. Treatment of myocardial infarction in a coronary care

unit: a two year experience with 250 patients. Am J Cardiol1967;20:457-464, 8. DeSanctis RW, Block P, Hotter AM. Tachyarrhythmias in myocardial infarction. Circulation 1972;45:681-702. 9. Engelman L. Stepwise logistic regression. In: Dixon WJ, ed. BMDP Statistical Software. Berkeley, California: University of California Press, 1981930-344. 10. 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. 11. Henning R. Swedish Cooperative CCU Study, part 2: the short term prognosis. Acta Med Stand 1975:198(suppl 586):1-35. 12. Lichstein E, Kuhn LA, Goldberg E, Mulvihill MN, Smith H, Chalmers TC. Diabetic treatment and primary ventricular fibrillation in acute myocardial infarction. Am J Cardiol 1976:38:100-102. 13. Tansey MJ, Opie LH, Kennelly BM. High mortality in obese women diabetics with acute myocardial infarction. Br Med J 1977;1;1624-1626. 14. Kereiakes DJ. Myocardial infarction in the diabetic patient. Chin Cardiol 1985;8:446-450. 15. Szklo M, Goldberg R, Kennedy HL, Tonascia JA. Survival in patients with nontransmural myocardial infarction: a population-based study. Am J Cardiol 1978;42:648-652. 16. Thanavaro S, Krone RJ, Kleiger RE, Province MA, Miller JP, deMello VR, Oliver GC. In-hospital prognosis of patients with first nontransmural and transmural infarctions. Circulation 1980;61;29-33. 17. Hatter AM, DcSanctis RW, Flynn T, Yeatman LA. Nontransmural myocar-

dial infarction: a comparison of hospital and late clinical course of patients with that of matched patients with transmural anterior and transmural inferior myocardial infarction. Am J Cardiol 1981;48:595-602. 18. Norris RM, Brandt PW, Caughey DE, Lee AJ, Scott PJ. A new coronary prognmtic index. Lnncet 1969;1:274. 19. Kitchin AH, Pocock SJ. Prognosis of patients with acute myocardial infarction admitted to a coronary care unit. I: Survival in hospital. Br Heart J 1977; 39:1163-1166. 20. Page DL, Caulfield JB, Kantor JA, DeSanctis RW, Sanders CA. Myocardial changes associated with cardiogenic shock. N Engl J Med 1971;285;133-137. 21. Opie LH. The glucose hypothesis: relation to acute myocardial ischemia. J Mel Cell Cardiol 1970;1:107-115. 22. Hearse DJ, Stewart DA, Chain EB. Diabetes and the survival and recovery of the anoxic mvocardium. J Mel Cell Cardiol 1975:7:397-415. 23. Fein FS,-Sonnenblick EH. Diabetic cardiomyopathy. Prog Cardiovasc Dis 1985:27:2S5-270. 24. Barrett-Connor E, Wingard DL. Sex differential in ischemic heart disease mortality in diabetics: a prospective population-based study. Am J Epidemiol 19833 18:489-496. 25. Walden CE, Knopp RH, Wahl PW, Beach KW, Strandness E. Sex differences in the effect of diabetes mellitus on lipoprotein triglyceride and cholesterol concentrations. N Engl J Med 1984;311:953-959. 26. Kannel WB, Hjortland M, Castelli WP. Role of diabetes in congestive heart failure: the Framingham Study. Am J Cardiol 1974;34:29-34.


acute myocardial infarction in diabetes mellitus and significance of congestive heart failure as a...

Documents