Sudden cardiac death after myocardial infarction in patients with type 2 diabetes M. Juhani Junttila, MD,*㛳* Petra Barthel, MD,†* Robert J. Myerburg, MD,㛳 Timo H. Mäkikallio, MD,* Axel Bauer, MD,† Kurt Ulm, PhD,‡ Antti Kiviniemi, PhD,§ Mikko Tulppo, PhD,§ Juha S. Perkiömäki, MD,* Georg Schmidt, MD,†† Heikki V Huikuri, MD*† From the *Institute of Clinical Medicine, Department of Internal Medicine, University of Oulu, Oulu, Finland, †Deutsches Herzzentrum und 1 Medizinische Klinik der Technischen Universität München, Munich, Germany, ‡Institut für Medizinische Statistik und Epidemiologie der Technischen Universität München, Munich, Germany, §Department of Exercise and Medical Physiology, Verve, Oulu, Finland, and 㛳Division of Cardiology, University of Miami Miller School of Medicine, Miami, Florida. BACKGROUND Type 2 diabetes mellitus is a well-established risk factor for atherosclerosis, but its contribution to sudden cardiac death (SCD) risk after myocardial infarction (MI) is not well defined. OBJECTIVE The purpose of this study was to compare the incidence and time-dependent risk of SCD in diabetic patients versus nondiabetic patients during 5-year follow-up after acute MI. METHODS A total of 3,276 patients were enrolled at the time of acute MI between 1996 and 2005. Mean age at entry was 60 ⫾ 11 years, and the cohort was followed until 2009. At entry into the study, diabetes was present in 629 (19.2%) patients. The primary endpoint was SCD, and the secondary endpoints were non-SCD and all-cause mortality.
An excess in the incidence of non-SCD began to appear among diabetic patients within the first 6 months of follow-up (P ⬍.001) but not in the incidence of SCD (P ⫽ .09). The excess in SCD among diabetic patients began to appear more than 6 months after the index event. CONCLUSION Patients with type 2 diabetes are at higher risk for SCD after MI than are nondiabetic patients. The incidence of SCD in post-MI type 2 diabetic patients with left ventricular ejection fraction ⬎35% is equal to that of nondiabetic patients with left ventricular ejection fraction ⬍35%. KEYWORDS Diabetes Mellitus; Myocardial infarction; Prognosis; Sudden death
RESULTS Among diabetic patients, the incidence of SCD was higher (5.9%) than in nondiabetic patients (1.7%), with a hazard ratio (HR) of 3.8 (95% confidence interval [CI] 2.4 –5.8; P ⬍.001) and adjusted HR of 2.3 (95% CI 1.4 –3.8; P ⬍.01). In diabetic patients with left ventricular ejection fraction ⬎35%, the incidence of SCD was nearly identical to that of nondiabetic patients with ventricular ejection fraction ⱕ35% (4.1% vs 4.9%; P ⫽ .48).
ABBREVIATIONS CI ⫽ confidence interval; HR ⫽ hazard ratio; ICD ⫽ implantable cardioverter-defibrillator; LVEF ⫽ left ventricular ejection fraction; MI ⫽ myocardial infarction; NYHA ⫽ New York Heart Association; SCD ⫽ sudden cardiac death
Introduction
factor for other chronic diseases, most notably cardiac and vascular disease. For example, the population subset of diabetic patients has a higher prevalence of coronary artery disease, and patients with type 2 diabetes without prior evidence of coronary artery disease are at increased risk for having cardiovascular events earlier than nondiabetic patients. Type 2 diabetic patients have an increased risk for cardiovascular mortality with or without evident coronary
Type 2 diabetes mellitus is a common disorder in developed areas of the world. It also is considered a prominent component of the currently emerging epidemic of chronic diseases in less developed areas of the world. Worldwide prevalence is estimated to reach nearly 220 million people by the year 2010.1 In addition to its implications as a primary disorder, type 2 diabetes mellitus is a powerful risk *Equal contributions as first author. †Equal contributions as senior author. This study was supported by research grants from the following nonprofit foundations: Fondation Leducq, Paris, France, to Drs. Junttila and Myerburg; Instrumentarium Science Foundation to Dr. Junttila; Orion-Farmos Research Foundation to Dr. Junttila; Finnish Funding Agency for Technology and Innovation, Helsinki, Finland, to Drs. Mäkikallio, Tulppo, Kiviniemi, and Huikuri; Research Council for Health, Academy of Finland, Helsinki, Finland, to Dr. Kiviniemi; Florida Heart Research Foundation, Miami, Florida, to Dr. Myerburg; Bundesministerium für Bildung, Wissenschaft, Forschung und
(Heart Rhythm 2010;7:1396 –1403) © 2010 Heart Rhythm Society. All rights reserved.
Technologie to Dr. Schmidt; Kommission für Klinische Forschung to Dr. Schmidt; and Deutsche Forschungsgemeinschaft to Dr. Schmidt. Dr. Junttila had full access to the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Address reprint requests and correspondence: Dr. M. Juhani Junttila, Division of Cardiology (D-39), University of Miami Miller School of Medicine, P.O. Box 016960, Miami, Florida 33101. E-mail address:
[email protected]. (Received July 12, 2010; accepted July 27, 2010.)
1547-5271/$ -see front matter © 2010 Heart Rhythm Society. All rights reserved.
doi:10.1016/j.hrthm.2010.07.031
Junttila et al
Sudden Death Among Diabetic Patients
artery disease at the time of diagnosis.2– 4 A number of studies have suggested that diabetes is a risk factor for sudden cardiac death (SCD) as well as for other mechanisms of cardiac mortality,5 but data regarding fatal and nonfatal cardiac event rates among diabetic subjects without prior myocardial infarction (MI) compared to nondiabetic subjects with prior MI are conflcting.6 – 8 In one study, the risk of future events in type 2 diabetic patients without MI was as high as in nondiabetic patients with prior MI.6 In another study, diabetic patients without MI were at lower risk for future events.7 Neither study provided data on measures of infarct size, history of heart failure, or SCD as a specific primary or secondary endpoint. SCD in the post-MI patient is of special interest because of the magnitude of risk, the challenge of individual risk profiling, and the potential for prevention in high-risk individuals.9 The previous observational studies comparing diabetic and nondiabetic patients with respect to death and cardiovascular morbidity after MI10 –13 or with respect to heart failure14 did not address the question of SCD risk. Only one of the MI studies10 and one heart failure study14 provided secondary analyses comparing SCD risk in diabetic and nondiabetic patients. A population-based observational study identified diabetes as a predictor of MI, nonSCD, and SCD, but the latter was not analyzed for SCD after MI because only 10% of the SCD cases had a prior MI.15 To our knowledge, no studies have focused on diabetes as a specific risk marker of SCD after MI. Furthermore, the duration of follow-up in prior studies of cardiovascular events in diabetic patients has been relatively short, and the majority of studies assessing the risk of cardiovascular mortality among diabetic patients were performed in 1990s, before the more recent advances in treatment of acute and chronic MI that improved the prognosis of affected patients. Because of current emphasis on the importance of identifying better clinical associations and biomarkers for SCD risk prediction, this study was designed to compare the contemporary incidence and temporal patterns of SCD in diabetic and nondiabetic patients during long-term follow-up after MI.
Methods Study population The study population consisted of enrollees in two prospective post-MI studies: the Multiple Risk Factor Analysis Trial (MRFAT) and the Improved Stratification of Autonomic Regulation for Risk Prediction postinfarction survey program (ISAR-Risk).16 The cohort consisted of a combined series of 3,276 acute MI patients (MRFAT n ⫽ 663; ISAR-Risk n ⫽ 2613) enrolled between January 1996 and January 2000 (MRFAT) and between January 2000 to March 2005 (ISAR-Risk). Part of the ISAR data was collected for a prospective evaluation of risk stratification measures among post-MI diabetic patients. Mean (⫾SD) age at entry was 60 ⫾ 11 years, and 77% of patients were males. Baseline characteristics and demographics of both groups are given in Table 4. Type 2 diabetes mellitus was identified prior to, or at the time of, entry into the study in 629 (19.2%)
1397 subjects. The diagnosis of type 2 diabetes was based upon World Health Organization (WHO) criteria, previously diagnosed type 2 diabetes, or plasma fasting glucose ⱖ7 mmol/L (126 mg/dL) or 2-hour plasma glucose ⱖ11.1 mmol/L (200 mg/dL).17 Patients with type 1 diabetes and patients with cardiac arrest during or prior to index MI were excluded from the study. The diagnosis of MI was based on the presence of at least two of three criteria from ICD-10: elevated troponin levels, ECG findings, and typical angina pectoris. The main objective of this cohort analysis was to compare the incidence and temporal distribution of SCD during long-term follow-up, comparing outcomes among type 2 diabetic and nondiabetic patients after MI, using acute and post-MI treatment strategies based on guidelines at the time of enrollment and during follow-up. Measurements and analyses of risk factor profiles, including heart rate variability/turbulence, and inclusion and exclusion criteria of this population have been previously described.8,10 –11,14 The observations in this report are based on a follow-up period that extended up to 5 years. This study was approved by the ethics committees and institutional review boards of both research sites, and all study subjects gave informed consent.
Endpoints The primary endpoint was SCD, and the secondary endpoints were non-SCD and all-cause mortality. The modes and causes of death were identified from hospital records and autopsy reports and from primary care physicians or those who had witnessed the death. Death certificates were obtained when needed. Independent endpoint committees at the University of Oulu and the University of Miami (MRFAT) and at Technischen Universität München (ISARRisk) adjudicated the mode and cause of all deaths, using the same criteria for SCD. Cardiac deaths were classified as sudden or nonsudden. Cardiac death was defined as sudden if it was (1) a witnessed death occurring within 60 minutes from the onset of new symptoms, unless a cause other than cardiac was obvious; (2) unwitnessed death (⬍24 hours) in the absence of preexisting, progressive circulatory failure or other cause of death; or (3) death during attempted resuscitation. All cardiac deaths that were not specified as sudden were considered non-SCD. Data from implantable cardioverter-defibrillator (ICD) implantation rates, programming, or ICD shocks were not included in the study.
Statistical analysis Baseline characteristics were compared by two-sided t-test and Chi-square test between type 2 diabetic patents and nondiabetic patients. Data were analyzed using SPSS 14.0 (SPSS Inc., Chicago, IL, USA). Kaplan-Meier survival curves were used to illustrate the incidence of SCD in follow-up time scale, and log rank analysis was performed to compare survival curves between groups. A subgroup analysis was performed with dichotomized left ventricular ejection fraction (LVEF) ⱕ35% or ⬎35% according to current guideline indications for primary prevention ICD
1398 Table 1
Heart Rhythm, Vol 7, No 10, October 2010 Baseline characteristics and demographic data of diabetic and nondiabetic patients
Variable Demographics Age (mean ⫾ SD; years) Gender (male %) Hypertension MI prior to index acute MI NYHA class (N ⫽ 1,525) I II III IV Thrombolysis Revascularization during hospitalization for index MI Left ventricular ejection fraction (%) SDNN (ms) Mean heart rate (bpm) Angiography: three-vessel disease (N ⫽ 2,851) Medication Beta-blocker therapy Angiotensin-converting enzyme inhibitor or angiotensin II receptor blocker therapy Antiplatelet therapy
Type 2 diabetic patients (n ⫽ 629)
Nondiabetic patients (n ⫽ 2,647)
P value
64 ⫾ 10 68% 74% 16%
59 ⫾ 11 78% 61% 10%
⬍.001 ⬍.001 ⬍.001 ⬍.001
68% 20% 8.3% 3.8% 10% 92% 49 ⫾ 12 88 ⫾ 37 68 ⫾ 11 42%
74% 16% 7.4% 2.3% 10% 92% 52 ⫾ 12 100 ⫾ 34 64 ⫾ 10 30%
95% 84%
95% 80%
.59 .04
99%
98%
.35
.13 .89 .71 ⬍.001 ⬍.001 ⬍.001 ⬍.001
NYHA class ⫽ New York Heart Association classification for heart failure/functional capacity; Revascularization ⫽ coronary artery bypass grafting or percutaneous coronary angioplasty during/shortly after baseline acute myocardial infarction (MI); SDNN ⫽ standard deviation of all normal R-R intervals; Thrombolysis (antithrombotic therapy) ⫽ aspirin, clopidogrel, and/or warfarin therapy.
therapy after acute MI.18 Univariate Cox proportional hazards regression analysis was used to estimate predictive power. In multivariate Cox regression analysis, all variables that differed between diabetic and nondiabetic patients at baseline and predefined variables, such as age and sex, were included in the model (Table 1). The results are given as hazard ratio (HR) and corresponding 95% confidence interval (CI) with P value. P ⬍.05 was considered significant. The primary analysis was carried out on the combined cohort. However, because of temporal, clinical, and interventional differences between the MRFAT and ISAR-Risk cohorts, separate analyses were performed in this study to determine whether the same findings regarding SCD in diabetic patients appeared in each group individually.
Results Study population Baseline characteristics of the study population, including comparisons between diabetic and nondiabetic patients, are given in Table 1. Diabetic patients were older (mean 64 ⫾ 10 years vs 59 ⫾ 11 years) and had a higher proportion of females (32% vs 22%), higher incidence of prior MI, higher incidence of hypertension, slightly lower LVEF (mean 49 ⫾ 12 vs 52 ⫾ 12), and higher incidence of three-vessel disease by angiography (42% vs 30%) compared to nondiabetic patients. However, New York Heart Association (NYHA) functional class at the time of discharge from the hospital after the index MI did not differ between the 313 diabetic and 1,212 nondiabetic patients for whom this information was available (Table 1).
The medications at baseline did not differ between the groups except for angiotensin-converting enzyme inhibitors or angiotensin II receptor blocking agents, which was more common among diabetic than nondiabetic patients. Percutaneous interventional or surgical revascularization procedures during the index MI hospitalization were performed in 92% of both diabetic and nondiabetic enrollees (Table 1). Standard deviation of all normal R-R intervals, a measure of heart rate variability, was significantly reduced, and mean heart rate was significantly higher in diabetic than nondiabetic patients.
Outcomes During mean follow-up of 4.0 years (median 5 ⫾ 1.4 [SD] years; 25th and 75th percentile ⫽ 2.9 and 5.0 years, respectively), SCD occurred in 5.9% of the diabetic subgroup compared to 1.7% of the nondiabetic subgroup (P ⬍.001; Table 2). Non-SCD occurred in 7.2% of diabetic patients compared to 2.8% of nondiabetic patients (P ⬍.001). Cumulative all-cause mortality was 21.6% among diabetic patients and 8.4% among nondiabetic patients (P ⬍.001). In univariate Cox regression analysis, diabetes was a significant predictor of the primary endpoint of SCD with HR of 3.8 (95% CI 2.4 –5.8; P ⬍.001). HR of all-cause mortality was 2.9 (95% CI 2.3–3.6; P ⬍.001) and that of non-SCD was 2.7 (95% CI 1.9 – 4.0; P ⬍.001). In multivariate analysis adjusted for gender, age, hypertension, prior MI, three-vessel disease, LVEF, measure of heart rate variability, mean heart rate, and use of angiotensin-converting enzyme inhibitors or angiotensin II receptor blocking agents, type 2 diabetes remained an independent predictor
Junttila et al
Sudden Death Among Diabetic Patients
1399
Table 2 Outcome events according to diabetes status and ejection fractions: sudden cardiac death, cardiac mortality, and all cause mortality Diabetic (type 2 diabetes mellitus) Primary Endpoint Sudden cardiac death Total* LVEF ⱕ35% LVEF ⬎35% Secondary Endpoints Non–sudden cardiac death Total* LVEF ⱕ35% LVEF ⬎35% All-cause mortality Total* LVEF ⱕ35% LVEF ⬎35%
Nondiabetic
Total
37/629 (5.9%) 15/95 (15.8%) 22/533 (4.1%)
46/2,647 (1.7%) 12/244 (4.9%) 34/2,400 (1.4%)
83/3,276 (2.5%) 27/339 (8.0%) 56/2,933 (1.9%)
45/629 (7.2%) 19/95 (20.0%) 25/533 (4.7%)
75/2,647 (2.8%) 32/244 (13.1%) 43/2,400 (1.8%)
120/3,276 (3.4%) 51/339 (15.0%) 68/2,933 (2.3%)
136/629 (21.6%) 48/95 (50.5%) 88/533 (16.5%)
222/2,647 (8.4%) 64/244 (26.2%) 158/2,400 (6.6%)
358/3,276 (10.9%) 112/339 (33.0%) 246/2,933 (8.4%)
Number of endpoints are grouped by diabetes status and LVEF ⱕ35% and LVEF ⬎35%. *Left ventricular ejection fraction (LVEF) available for 628/629 diabetic patients and 2,644/2,647 nondiabetic patients (⬎99% of cohort).
of SCD with HR of 2.3 (95% CI 1.4 –3.8; P ⬍.01). Diabetes also remained a marginal independent predictor of non-SCD with HR of 1.6 (95% CI 1.04 –2.5; P ⫽ .03). Detailed results of Cox regression analyses are given in Table 3.
Temporal patterns Kaplan-Meier analysis revealed that the frequency of SCD events did not begin to increase until 6 months after the index event and that separation of the curves for SCD risk in diabetic and nondiabetic patients began at the same time (6-month follow-up point log rank P ⫽ .09; Figure 1). This pattern persisted for at least 4 years after the acute MI. In Table 3
contrast, the incidence of non-SCD was higher than the incidence of SCD during the first 6 months (log rank P ⬍.001) after the index event, and the diabetic and nondiabetic curves began to diverge at that earlier time and had started to run parallel by 2 years. The curves reveal that mortality risk, and especially SCD risk, was both greater in magnitude and expressed earlier after MI in the diabetic subgroup.
Left ventricular ejection fraction In a subgroup analysis comparing patients with LVEF ⱕ35% versus those with LVEF ⬎35%, the incidence of
Univariate and multivariate predictors of sudden cardiac death, non–sudden cardiac death, and all-cause mortality
Variable Univariate Diabetes Hypertension Prior MI Three-vessel disease LVEF SDNN Mean heart rate from 24-hour Holter recording Angiotensin-converting enzyme inhibitor or angiotensin II receptor blocker Multivariate Diabetes Hypertension Prior MI Three-vessel disease LVEF SDNN Mean heart rate from 24-hour Holter recording Angiotensin-converting enzyme inhibitor or angiotensin II receptor blocker
SCD
Non-SCD
All-cause mortality
3.8‡ 1.2 2.4† 3.5‡ 0.94‡ 0.99† 1.03‡ 0.65
(2.4–5.8) (0.74–1.8) (1.4–4.0) (2.1–5.7) (0.92–0.95) (0.98–0.99) (1.01–1.05) (0.41–1.05)
2.7‡ 1.2 4.1‡ 2.2‡ 0.93‡ 0.99† 1.04‡ 1.3
(1.9–4.0) (0.84–1.8) (2.8–5.9) (1.5–3.3) (0.91–0.94) (0.98–0.99) (1.02–1.06) (0.79–2.1)
2.9‡ 1.2 2.8‡ 2.3‡ 0.94‡ 0.99‡ 1.03‡ 0.94
(2.3–3.6) (0.99–1.5) (2.2–3.5) (1.8–3.0) (0.92–0.95) (0.98–0.99) (1.02–1.04) (0.73–1.2)
2.3† 1.1 1.5 2.5† 0.95‡ 1.00 1.03* 0.31‡
(1.4–3.8) (0.65–1.9) (0.84–2.7) (1.5–4.3) (0.93–0.97) (0.99–1.01) (1.01–1.05) (0.17–0.55)
1.6* 1.01 1.7* 1.1 0.94‡ 1.00 1.04‡ 1.03
(1.04–2.5) (0.64–1.6) (1.06–2.7) (0.75–1.8) (0.92–0.96) (0.99–1.01) (1.02–1.06) (0.51–2.1)
1.9‡ 1.01 1.6† 1.4* 0.96‡ 1.00 1.03‡ 0.72
(1.5–2.5) (0.78–1.3) (1.2–2.1) (1.1–1.8) (0.95–0.97) (0.99–1.01) (1.01–1.04) (0.50–1.03)
Values are given as hazard ratio (95% confidence interval). Multivariate analysis also adjusted for gender and age. LVEF ⫽ left ventricular ejection fraction (continuous variable); MI ⫽ myocardial infarction; SCD ⫽ sudden cardiac death; SDNN ⫽ standard deviation of all normal R-R intervals (ms). *P ⬍.05; †P ⬍.01, ‡P ⬍.001.
1400
Heart Rhythm, Vol 7, No 10, October 2010
Separate cohort analyses The MRFAT and ISAR-Risk subgroups differed in a number of baseline variables, such as frequency of acute interventions during index MI (76% vs 96%; P ⬍.001), LVEF (45 ⫾ 9 vs 53 ⫾ 13; P ⬍.001), and time of enrollment (1996 –1999 vs 1996 –2005). Baseline differences that were statistically significant are listed in Table 4. Separate analyses of the two cohorts revealed that these baseline differences were associated with differences in absolute numbers of events, but the general relationships between event rates in type 2 diabetic patients and nondiabetic patients were maintained. In the MRFAT population, diabetic patients had higher rate of SCD (9.5%) compared to nondiabetic patients (4.5%) with adjusted HR of 2.2 (95% CI 1.0 – 4.9; P ⫽ .02). In the ISAR-Risk population, diabetic patients also had a higher rate of SCD (4.8%) compared to nondiabetic patients (1.5%) with adjusted HR of 2.4 (95% CI 1.4 – 4.2; P ⫽ .001).
Discussion
Figure 1 Incidence of non–sudden cardiac death (non-SCD) and sudden cardiac death (SCD) in type 2 diabetic and nondiabetic patients after acute myocardial infarction. A: Kaplan-Meier curve for non-SCD shows the significant difference between diabetic and nondiabetic patients. The difference becomes evident early after myocardial infarction. B: Curve for SCD illustrates the significant difference of SCD incidence between diabetic and nondiabetic patients. The slope of the accelerated portion of the delayed SCD risk curve is much steeper for diabetic patients than for nondiabetic patients.
SCD among diabetic patients with LVEF ⬎35% was nearly identical to that of nondiabetic patients with LVEF ⱕ35% (4.1% vs 4.9%; log rank P ⫽ .48). In contrast, the incidence of non-SCD was significantly lower among diabetic patients with LVEF ⬎35% than among nondiabetic patients with LVEF ⱕ35% (4.7% vs 13.1%; log rank P ⬍.001). The incidence of SCD among diabetic patients with LVEF ⱕ35% was especially high (15.8%, log rank P ⬍.001 compared to other groups). KaplanMeier curves illustrating these SCD incidences are presented in Figure 2.
This study demonstrates a higher incidence of SCD among type 2 diabetic patients during long-term follow-up after MI compared to nondiabetic patients (5.9% vs 1.7%). NonSCD also was higher among diabetic than nondiabetic patients (7.2% vs 2.8%). Increased risk of SCD in diabetic patients remained significant after adjustments for baseline differences between diabetic and nondiabetic patients at the time of acute MI, but the difference in non-SCD between diabetic and nondiabetic patients was less robust after adjustment for these variables. Notably, the prevalence and severity of heart failure as reflected in NYHA functional classification did not differ between diabetic and nondiabetic patients (Table 1), despite baseline differences in variables related to extent of disease, such as prevalence of three-vessel coronary artery disease, history of MI prior to the index MI, and a clinically small but statistically significant difference in post-MI LVEF. Given the role of heart failure as a predictor of SCD after MI,9 this observation highlights the potential importance of type 2 diabetes as an independent modulator of risk. Even differences in the measure of autonomic function, such as heart rate variability, did not have an effect on diabetes as an independent risk marker of SCD. In an analysis of prespecified subgroups, the incidence of SCD among post-MI type 2 diabetic patients with LVEF greater than 35% was almost identical to the incidence among nondiabetic patients with LVEF less than 35% (Figure 2). Furthermore, a very high SCD incidence (15.8%) was observed with the combination of LVEF ⬍35% after MI and type 2 diabetes. These data suggest that type 2 diabetes modulates risk of SCD in post-MI patients with LVEF ⬍35% and in those with LVEF ⬎35%. The observation of equivalent risk between diabetic patients with LVEF ⬎35% and nondiabetic patients with LVEF ⬍35%, as well as a higher risk in diabetic than in nondiabetic patients with LVEF ⬍35%, may have implications for interpretation of evidence- and judgment-based criteria for
Junttila et al
Sudden Death Among Diabetic Patients
1401
Figure 2 Kaplan-Meier curves illustrate the difference in incidence of non–sudden cardiac death (Non-SCD; A) and sudden cardiac death (SCD; B) for diabetic patients with left ventricular ejection fraction (EF) ⬎35% and nondiabetic patients in the same EF group. The curves also illustrate the identical incidence of SCD between diabetic patients with EF ⬎35% and nondiabetic patients with EF ⱕ35% and the significantly higher incidence of non-SCD in the latter group compared to diabetic patients with EF ⬎35%. Diabetic patients with EF ⱕ35% have an extraordinary high incidence of sudden cardiac death after myocardial infarction. DM ⫽ type 2 diabetes mellitus.
ICD therapy after MI.19 However, this observation should not translate to an alteration of ICD indications without validation of an added survival benefit of such therapy in diabetic patients given the high non-SCD risk. The results highlight the suggested need for further prospective studies to determine the role of the diabetic state for better profiling of SCD risk generally9 and for consideration of future modifications of ICD guidelines for post-MI patients specifically. In the meantime, clinical efforts should continue to focus on prevention of recur-
rent ischemic events in diabetic patients, which in fact may be a causal link between the increased risk of SCD and type 2 diabetes. The separate cohort analyses, performed because of baseline differences, revealed quantitative differences in outcomes between the two cohorts, but the pattern of excesses in SCD among diabetics after MI remained in both cohorts, as did the patterns related to LVEF subgroups. These observations suggest that the greater use of primary percutaneous interventions during acute MI in the ISAR-
1402 Table 4
Heart Rhythm, Vol 7, No 10, October 2010 Baseline characteristics/demographics, interventions, and outcome data for MRFAT and ISAR-Risk cohorts
Baseline Characteristic Age (years) Hypertension Diabetes MI prior to index MI Intervention Revascularization during hospitalization for index MI Beta-blocker therapy Angiotensin-converting enzyme inhibitor or angiotensin II receptor blocker therapy Antiplatelet therapy Outcome SCD Non-SCD All-cause mortality
MRFAT (n ⫽ 663)
ISAR-Risk (n ⫽ 2,613)
P value
62 ⫾ 10 325 (49%) 148 (22%) 60 (9%)
60 ⫾ 11 1741 (67%) 481 (18%) 312 (12%)
⬍.001 ⬍.001 .02 .04
503 (76%) 97% 39%
2509 (96%) 94% 91%
⬍.001 .01 ⬍.001
99% DM/Non-DM 9.5%/4.5% 11.5%/3.9% 36.5%/13.6%
98% DM/Non-DM 4.8%/1.5% 5.8%/2.6% 17.3%/7.1%
.01 MRFAT/ISAR-Risk ⬍.001/⬍.001 ⬍.001/⬍.001 ⬍.001/⬍.001
Baseline characteristics with significant differences between groups are listed. DM ⫽ type 2 diabetes mellitus; MI ⫽ myocardial infarction; SCD ⫽ sudden cardiac death.
Risk cohort did not negate the impact of type 2 diabetes as a modulator of SCD risk after MI. The distributions of SCD over time are complex and provocative. Previous data demonstrated that the risk of SCD after acute MI is high during the first 6 months after the index event,20 whereas other data indicated that SCD risk is delayed for 2 to 4 years and beyond,12,21–24 the combination suggesting a bimodal pattern of risk. Our data reaffirm delayed risk of SCD after MI for both diabetic and nondiabetic patients but also suggest that the diabetic subgroup expresses not only as higher risk but also concentrates SCD earlier than in nondiabetic patients. Figure 1 shows that the slope of the accelerated portion of the delayed SCD risk curve (6 months and beyond) is much steeper for diabetic patients than is the corresponding curve for nondiabetic patients. Earlier studies described increased all-cause mortality and cardiovascular mortality among type 2 diabetic patients with MI.4 – 6 Most of these previous studies were performed between 1980 and 1995. Interventional and pharmacologic treatment strategies have evolved dramatically from those times, and cardiovascular mortality in the current treatment era among type 2 diabetic MI patients has not been thoroughly defined. One post-MI study alluded to a higher incidence of SCD,10 another study suggested that diabetes increases risk of SCD in heart failure patients,14 and one large population study found an association between diabetes and SCD risk.15 The latter differs from our study in that only a small proportion of SCDs (⬃10%) were post-MI events. To our knowledge, this is the first study that focused specifically on the comparative long-term risk of SCD in type 2 diabetic patients after MI. The cumulative 5-year incidence of post-MI SCD was somewhat lower than expected from prior studies, particularly for the nondiabetic patients. This may be due in part to the relatively high LVEF after MI, the influence of the high rate of interventional procedures in the era studied, or both. Nonetheless, the incidence of SCD increased between 6
months and 4 years after acute MI among type 2 diabetic patients. This observation differs from, and is likely complementary to, the concept of a phase of early vulnerability to SCD during the convalescent phase after acute MI. However, we did observe excess mortality due to non-SCD during the first few months after the index event. The reasons for these specific temporal patterns of SCD and non-SCD among the diabetic patients are speculative. Among the diabetic patients, a high 5-year all-cause mortality rate (21.6% vs 8.4% in nondiabetic patients; Table 2) was due in large part to noncardiac causes. Among diabetic patients with LVEF ⱕ35%, the 5-year mortality reached 50.5% compared to 26.2% in nondiabetic patients, figures proportional to the non-SCD and SCD risks in the low LVEF subgroup (Table 2). The latter suggests a significant contribution of left ventricular dysfunction to all-cause mortality risk in the low LVEF subgroup. Diabetic patients with severely impaired left ventricular function have a high risk of dying due to progressive heart failure early after MI, as do nondiabetic patients. Those who do not die of heart failure early after MI remain at increased risk for experiencing SCD late after the event. The latter may be due partly to early protection against SCD attributable to beta-blocker therapy,12 which is overridden later by either ventricular remodeling or increased rate of recurrent coronary events. Recurrent ischemic events in the scarred myocardium have long been recognized as an important factor predisposing to the onset of fatal arrhythmias.25–27 An alternative and/or additive mechanism is suggested by the reported association between diabetes and QT-interval duration in SCD victims observed in a community-based population study of the causes and mechanisms of SCD.28
Study limitations Some limitations affect the interpretation of the study findings. Although there was no difference between the diabetic and nondiabetic subgroups with regard to a history of heart failure or heart failure at the time of acute MI, we do not
Junttila et al
Sudden Death Among Diabetic Patients
have data on interim incident heart failure during follow-up. A similar limitation applies to interim MI, revascularization procedures, and changes in LVEF over time. Uneven distribution of such time-dependent variables between diabetic and nondiabetic patients might have affected the differences in SCD risk. Due to the observational nature of this death certificate– based mortality study, data on a number of interim risk factors that may be more prevalent in diabetics could not be evaluated, such as development of renal or autonomic dysfunction, changing patterns of hemoglobin A1c levels, and transition to insulin therapy. In addition, we could not identify the number of SCD that were due to mechanisms other than tachyarrhythmias and thus would not be prevented by an ICD.
Conclusion Patients with type 2 diabetes are at higher risk for SCD after MI than are nondiabetic patients. The incidence of SCD in post-MI type 2 diabetic patients with LVEF ⬎35% is equal to that of nondiabetic patients with LVEF ⬍35%. Further prospective information on post-MI diabetic patients is needed to evaluate indications for, and efficacy of, ICDs and other therapies for this higher-risk subgroup.
References 1. Zimmet P. Globalization, coca-colonization and the chronic disease epidemic: can the doomsday scenario be averted? J Intern Med 2000;247:301–310. 2. Kannel WB, McGee DL. Diabetes and cardiovascular risk factors: the Framingham study. Circulation 1979;59:8 –13. 3. Stamler J, Vaccaro O, Neaton JD, Wentworth D. Diabetes, other risk factors, and 12-yr cardiovascular mortality for men screened in the multiple risk factor intervention trial. Diabetes Care 1993;16:434 – 444. 4. Morrish NJ, Wang SL, Stevens LK, Fuller JH, Keen H. Mortality and causes of death in the WHO multinational study of vascular disease in diabetes. Diabetologia 2001;44(Suppl 2):S14 –S21. 5. Siscovick DS, Sotoodehnia N, Rea TD, Ragunathan TE, Jouven X, Lemaitre RN. Type 2 diabetes and the risk of sudden cardiac arrest in the community. Rev Endocr Metab Disord 2010;11:53–59. 6. 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. 7. Evans JM, Wang J, Morris AD. Comparison of cardiovascular risk between patients with type 2 diabetes and those who had had a myocardial infarction: Cross sectional and cohort studies. BMJ 2002;324:939 –942. 8. Stephens JW, Ambler G, Vallance P, Betteridge DJ, Humphries SE, Hurel SJ. Cardiovascular risk and diabetes. are the methods of risk prediction satisfactory? Eur J Cardiovasc Prev Rehabil 2004;11:521–528. 9. Myerburg RJ. Implantable cardioverter-defibrillators after myocardial infarction. N Engl J Med 2008;359:2245–2253. 10. Makikallio TH, Barthel P, Schneider R, et al. Prediction of sudden cardiac death after acute myocardial infarction: role of Holter monitoring in the modern treatment era. Eur Heart J 2005;26:762–769.
1403 11. Exner DV, Kavanagh KM, Slawnych MP, et al. Noninvasive risk assessment early after a myocardial infarction the REFINE study. J Am Coll Cardiol 2007;50:2275–2284. 12. Huikuri HV, Tapanainen JM, Lindgren K, et al. Prediction of sudden cardiac death after myocardial infarction in the beta-blocking era. J Am Coll Cardiol 2003;42:652– 658. 13. Tapanainen JM, Still AM, Airaksinen KE, Huikuri HV. Prognostic significance of risk stratifiers of mortality, including T wave alternans, after acute myocardial infarction: results of a prospective follow-up study. J Cardiovasc Electrophysiol 2001;12:645– 652. 14. MacDonald MR, Petrie MC, Varyani F, et al. Impact of diabetes on outcomes in patients with low and preserved ejection fraction heart failure: an analysis of the Candesartan in Heart failure: Assessment of Reduction in Mortality and morbidity (CHARM) programme. Eur Heart J 2008;29:1377–1385. 15. Kucharska-Newton AM, Couper DJ, Pankow JS, et al. Diabetes and the risk of sudden cardiac death, the Atherosclerosis Risk in Communities study. Acta Diabet 2009;[Epub October 24, 2009]. 16. Bauer A, Barthel P, Schneider R, et al. Improved Stratification of Autonomic Regulation for risk prediction in post-infarction patients with preserved left ventricular function (ISAR-Risk). Eur Heart J 2009;30:576 –583. 17. Alberti KG, Zimmet PZ. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabet Med 1998;15:539 – 553. 18. Zipes DP, Camm AJ, Borggrefe M, et al. ACC/AHA/ESC 2006 guidelines for management of patients with ventricular arrhythmias and the prevention of sudden cardiac death: a report of the American College of Cardiology/American Heart Association task force and the European Society of Cardiology committee for practice guidelines (writing committee to develop guidelines for management of patients with ventricular arrhythmias and the prevention of sudden cardiac death): Developed in collaboration with the European Heart Rhythm Association and the Heart Rhythm Society. Circulation 2006;114:e385– e484. 19. Myerburg RJ, Reddy V, Castellanos A. Indications for implantable cardioverterdefibrillators based on evidence and judgment. J Am Coll Cardiol 2009; 54: 747–763. 20. Solomon SD, Zelenkofske S, McMurray JJ, et al. Sudden death in patients with myocardial infarction and left ventricular dysfunction, heart failure, or both. N Engl J Med 2005;352:2581–2588. 21. Moss AJ, Zareba W, Hall WJ, et al. Prophylactic implantation of a defibrillator in patients with myocardial infarction and reduced ejection fraction. N Engl J Med 2002;346:877– 883. 22. Bardy GH, Lee KL, Mark DB, et al. Amiodarone or an implantable cardioverterdefibrillator for congestive heart failure. N Engl J Med 2005;352:225–237. 23. Wilber DJ, Zareba W, Hall WJ, et al. Time dependence of mortality risk and defibrillator benefit after myocardial infarction. Circulation 2004;109:1082– 1084. 24. Goldenberg I, Vyas AK, Hall WJ, et al. Risk stratification for primary implantation of a cardioverter-defibrillator in patients with ischemic left ventricular dysfunction. J Am Coll Cardiol 2008;51:288 –296. 25. Huikuri HV, Castellanos A, Myerburg RJ. Sudden death due to cardiac arrhythmias. N Engl J Med 2001;345:1473–1482. 26. Furukawa T, Moroe K, Mayrovitz HN, Sampsell R, Furukawa N, Myerburg RJ. Arrhythmogenic effects of graded coronary blood flow reductions superimposed on prior myocardial infarction in dogs. Circulation 1991;84:368 –377. 27. Myerburg RJ, Kessler KM, Castellanos A. Sudden cardiac death: Epidemiology, transient risk, and intervention assessment. Ann Intern Med 1993;119:1187– 1197. 28. Chugh SS, Reinier K, Singh T, et al. Determinants of prolonged QT interval and their contribution to sudden death risk in coronary artery disease: the Oregon Sudden Unexpected Death Study. Circulation 2009;119:663– 670.
