[Skip to Content]
Access to paid content on this site is currently suspended due to excessive activity being detected from your IP address 34.204.52.4. Please contact the publisher to request reinstatement.
[Skip to Content Landing]
Figure.
Cumulative Incidence Rates of Heart Failure According to Coronary Artery Disease Categories
Cumulative Incidence Rates of Heart Failure According to Coronary Artery Disease Categories

Death and recurrent myocardial infarction (MI) were treated as competing events. Follow-up began at the time of the index MI and was truncated at 6.7 years (the mean follow-up duration in this analysis).

Table 1.  
Characteristics of Patients With Incident MI
Characteristics of Patients With Incident MI
Table 2.  
Heart Failure Incidence
Heart Failure Incidence
1.
Wald  DS, Morris  JK, Wald  NJ,  et al; PRAMI Investigators.  Randomized trial of preventive angioplasty in myocardial infarction.  N Engl J Med. 2013;369(12):1115-1123.PubMedGoogle ScholarCrossref
2.
Mauri  L.  Nonculprit lesions—or guilty by association.  N Engl J Med. 2013;369(12):1166-1167.PubMedGoogle ScholarCrossref
3.
El-Hayek  GE, Gershlick  AH, Hong  MK,  et al.  Meta-analysis of randomized controlled trials comparing multivessel versus culprit-only revascularization for patients with ST-segment elevation myocardial infarction and multivessel disease undergoing primary percutaneous coronary intervention.  Am J Cardiol. 2015;115(11):1481-1486.PubMedGoogle ScholarCrossref
4.
Gerber  Y, Weston  SA, Berardi  C,  et al.  Contemporary trends in heart failure with reduced and preserved ejection fraction after myocardial infarction: a community study.  Am J Epidemiol. 2013;178(8):1272-1280.PubMedGoogle ScholarCrossref
5.
McManus  DD, Chinali  M, Saczynski  JS,  et al.  30-Year trends in heart failure in patients hospitalized with acute myocardial infarction.  Am J Cardiol. 2011;107(3):353-359.PubMedGoogle ScholarCrossref
6.
Chen  J, Hsieh  AF, Dharmarajan  K, Masoudi  FA, Krumholz  HM.  National trends in heart failure hospitalization after acute myocardial infarction for Medicare beneficiaries: 1998-2010.  Circulation. 2013;128(24):2577-2584.PubMedGoogle ScholarCrossref
7.
Sutton  MG, Sharpe  N.  Left ventricular remodeling after myocardial infarction: pathophysiology and therapy.  Circulation. 2000;101(25):2981-2988.PubMedGoogle ScholarCrossref
8.
Wijns  W, Vatner  SF, Camici  PG.  Hibernating myocardium.  N Engl J Med. 1998;339(3):173-181.PubMedGoogle ScholarCrossref
9.
Gheorghiade  M, Bonow  RO.  Chronic heart failure in the United States: a manifestation of coronary artery disease.  Circulation. 1998;97(3):282-289.PubMedGoogle ScholarCrossref
10.
Velagaleti  RS, Vasan  RS.  Heart failure in the twenty-first century: is it a coronary artery disease or hypertension problem?  Cardiol Clin. 2007;25(4):487-495.PubMedGoogle ScholarCrossref
11.
Hwang  SJ, Melenovsky  V, Borlaug  BA.  Implications of coronary artery disease in heart failure with preserved ejection fraction.  J Am Coll Cardiol. 2014;63(25, pt A):2817-2827.PubMedGoogle ScholarCrossref
12.
Owan  TE, Hodge  DO, Herges  RM, Jacobsen  SJ, Roger  VL, Redfield  MM.  Trends in prevalence and outcome of heart failure with preserved ejection fraction.  N Engl J Med. 2006;355(3):251-259.PubMedGoogle ScholarCrossref
13.
Lam  CS, Donal  E, Kraigher-Krainer  E, Vasan  RS.  Epidemiology and clinical course of heart failure with preserved ejection fraction.  Eur J Heart Fail. 2011;13(1):18-28.PubMedGoogle ScholarCrossref
14.
Bhatia  RS, Tu  JV, Lee  DS,  et al.  Outcome of heart failure with preserved ejection fraction in a population-based study.  N Engl J Med. 2006;355(3):260-269.PubMedGoogle ScholarCrossref
15.
Tribouilloy  C, Rusinaru  D, Mahjoub  H,  et al.  Prognosis of heart failure with preserved ejection fraction: a 5 year prospective population-based study.  Eur Heart J. 2008;29(3):339-347.PubMedGoogle ScholarCrossref
16.
United States Census. State & County QuickFacts. http://quickfacts.census.gov/qfd/states/27/27109.html. Accessed October 23, 2015.
17.
Melton  LJ  III.  History of the Rochester Epidemiology Project.  Mayo Clin Proc. 1996;71(3):266-274.PubMedGoogle ScholarCrossref
18.
St Sauver  JL, Grossardt  BR, Yawn  BP, Melton  LJ  III, Rocca  WA.  Use of a medical records linkage system to enumerate a dynamic population over time: the Rochester Epidemiology Project.  Am J Epidemiol. 2011;173(9):1059-1068.PubMedGoogle ScholarCrossref
19.
Roger  VL, Weston  SA, Gerber  Y,  et al.  Trends in incidence, severity, and outcome of hospitalized myocardial infarction.  Circulation. 2010;121(7):863-869.PubMedGoogle ScholarCrossref
20.
Scanlon  PJ, Faxon  DP, Audet  AM,  et al.  ACC/AHA guidelines for coronary angiography: a report of the American College of Cardiology/American Heart Association Task Force on practice guidelines (Committee on Coronary Angiography): developed in collaboration with the Society for Cardiac Angiography and Interventions.  J Am Coll Cardiol. 1999;33(6):1756-1824.PubMedGoogle ScholarCrossref
21.
Charlson  ME, Pompei  P, Ales  KL, MacKenzie  CR.  A new method of classifying prognostic comorbidity in longitudinal studies: development and validation.  J Chronic Dis. 1987;40(5):373-383.PubMedGoogle ScholarCrossref
22.
Levey  AS, Coresh  J, Greene  T,  et al; Chronic Kidney Disease Epidemiology Collaboration.  Using standardized serum creatinine values in the Modification of Diet in Renal Disease study equation for estimating glomerular filtration rate.  Ann Intern Med. 2006;145(4):247-254.PubMedGoogle ScholarCrossref
23.
Killip  T  III, Kimball  JT.  Treatment of myocardial infarction in a coronary care unit. A two year experience with 250 patients.  Am J Cardiol. 1967;20(4):457-464.PubMedGoogle ScholarCrossref
24.
McKee  PA, Castelli  WP, McNamara  PM, Kannel  WB.  The natural history of congestive heart failure: the Framingham study.  N Engl J Med. 1971;285(26):1441-1446.PubMedGoogle ScholarCrossref
25.
Roger  VL, Weston  SA, Redfield  MM,  et al.  Trends in heart failure incidence and survival in a community-based population.  JAMA. 2004;292(3):344-350.PubMedGoogle ScholarCrossref
26.
Dunlay  SM, Roger  VL, Weston  SA, Jiang  R, Redfield  MM.  Longitudinal changes in ejection fraction in heart failure patients with preserved and reduced ejection fraction.  Circ Heart Fail. 2012;5(6):720-726.PubMedGoogle ScholarCrossref
27.
Yancy  CW, Jessup  M, Bozkurt  B,  et al; Writing Committee Members; American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines.  2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association Task Force on practice guidelines.  Circulation. 2013;128(16):e240-e327.PubMedGoogle ScholarCrossref
28.
Cox  DR.  Regression analysis and life table.  J R Stat Soc B. 1972;34:187-222.Google Scholar
29.
Austin  PC.  The use of propensity score methods with survival or time-to-event outcomes: reporting measures of effect similar to those used in randomized experiments.  Stat Med. 2014;33(7):1242-1258.PubMedGoogle ScholarCrossref
30.
Hernán  MA, Robins  JM.  Causal Inference. Boca Raton, FL: Chapman & Hall/CRC; 2016, forthcoming.
31.
Wolbers  M, Koller  MT, Witteman  JC, Steyerberg  EW.  Prognostic models with competing risks: methods and application to coronary risk prediction.  Epidemiology. 2009;20(4):555-561.PubMedGoogle ScholarCrossref
32.
Fine  JP, Gray  RJ.  A proportional hazards model for the subdistribution of a competing risk.  J Am Stat Assoc. 1999;94(446):496-509.Google ScholarCrossref
33.
Zhang  X, Zhang  MJ.  SAS macros for estimation of direct adjusted cumulative incidence curves under proportional subdistribution hazards models.  Comput Methods Programs Biomed. 2011;101(1):87-93.PubMedGoogle ScholarCrossref
34.
Gerber  Y, Jaffe  AS, Weston  SA, Jiang  R, Roger  VL.  Prognostic value of cardiac troponin T after myocardial infarction: a contemporary community experience.  Mayo Clin Proc. 2012;87(3):247-254.PubMedGoogle ScholarCrossref
35.
O’Connor  CM, Hathaway  WR, Bates  ER,  et al.  Clinical characteristics and long-term outcome of patients in whom congestive heart failure develops after thrombolytic therapy for acute myocardial infarction: development of a predictive model.  Am Heart J. 1997;133(6):663-673.PubMedGoogle ScholarCrossref
36.
Janardhanan  R, Kenchaiah  S, Velazquez  EJ,  et al; VALIANT Investigators.  Extent of coronary artery disease as a predictor of outcomes in acute myocardial infarction complicated by heart failure, left ventricular dysfunction, or both.  Am Heart J. 2006;152(1):183-189.PubMedGoogle ScholarCrossref
37.
Steg  PG, López-Sendón  J, Lopez de Sa  E,  et al; GRACE Investigators.  External validity of clinical trials in acute myocardial infarction.  Arch Intern Med. 2007;167(1):68-73.PubMedGoogle ScholarCrossref
38.
Yeh  RW, Sidney  S, Chandra  M, Sorel  M, Selby  JV, Go  AS.  Population trends in the incidence and outcomes of acute myocardial infarction.  N Engl J Med. 2010;362(23):2155-2165.PubMedGoogle ScholarCrossref
39.
Gerber  Y, Weston  SA, Jiang  R, Roger  VL.  The changing epidemiology of myocardial infarction in Olmsted County, Minnesota, 1995-2012.  Am J Med. 2015;128(2):144-151.PubMedGoogle ScholarCrossref
40.
Felker  GM, Shaw  LK, O’Connor  CM.  A standardized definition of ischemic cardiomyopathy for use in clinical research.  J Am Coll Cardiol. 2002;39(2):210-218.PubMedGoogle ScholarCrossref
41.
Bolognese  L, Neskovic  AN, Parodi  G,  et al.  Left ventricular remodeling after primary coronary angioplasty: patterns of left ventricular dilation and long-term prognostic implications.  Circulation. 2002;106(18):2351-2357.PubMedGoogle ScholarCrossref
Original Investigation
May 2016

Atherosclerotic Burden and Heart Failure After Myocardial Infarction

Author Affiliations
  • 1Division of Cardiovascular Diseases, Department of Health Sciences Research, Mayo Clinic, Rochester, Minnesota
  • 2Department of Epidemiology and Preventive Medicine, School of Public Health, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel
 

Copyright 2016 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.

JAMA Cardiol. 2016;1(2):156-162. doi:10.1001/jamacardio.2016.0074
Abstract

Importance  Whether the extent of coronary artery disease (CAD) is associated with the occurrence of heart failure (HF) after myocardial infarction (MI) is not known. Furthermore, whether this association might differ by HF type according to preserved or reduced ejection fraction (EF) has yet to be determined.

Objectives  To evaluate in a community cohort of patients with incident (first-ever) MI the association of angiographic CAD with subsequent HF and to examine the prognostic role of CAD according to HF subtypes: HF with reduced EF and HF with preserved EF.

Design, Setting, and Participants  A population-based cohort study was conducted in 1922 residents of Olmsted County, Minnesota, with incident MI diagnosed between January 1, 1990, and December 31, 2010, and no prior HF; study participants were followed up through March 31, 2013. The extent of angiographic CAD was determined at baseline and categorized according to the number of major epicardial coronary arteries with 50% or more lumen diameter obstruction.

Main Outcomes and Measures  The primary end point was time to incident HF. The primary exposure variable was the extent of CAD as expressed by the number of major coronary arteries with significant obstruction (0-, 1-, 2-, or 3-vessel disease) obtained from coronary angiograms performed no more than 1 day after the MI. Heart failure was ascertained by the Framingham criteria and classified by type according to EF (50% cutoff).

Results  Of the 1922 participants, 1258 (65.4%) were men (mean [SD] age, 64 [13] years). During a mean follow-up period of 6.7 (5.9) years, 588 patients (30.6%) developed HF. With death and recurrent MI modeled as competing risks, the cumulative incidence rates of post-MI HF among patients with 0 or 1, 2, and 3 diseased vessels were 10.7%, 14.6%, and 23.0% at 30 days; and 14.7%, 20.6%, and 29.8% at 5 years, respectively (P < .001 for trend). After adjustment for clinical characteristics in a Cox proportional hazards regression model, the hazard ratios (95% CIs) for HF were 1.25 (0.99-1.59) and 1.75 (1.40-2.20) in patients with 2 and 3 vessels vs 0 or 1 occluded vessel, respectively (P < .001 for trend). The increased risk with a greater number of occluded vessels was independent of the occurrence of a recurrent MI and did not differ appreciably by HF type.

Conclusions and Relevance  The extent of angiographic CAD is an indicator of post-MI HF regardless of HF type and independent of recurrent MI. These data underscore the need to further investigate the processes taking place in the transition from myocardial injury to HF.

Introduction

Several recent publications1-3 have drawn attention to the extent of coronary disease (CAD) as a therapeutic target in acute myocardial infarction (MI) beyond the treatment of the culprit lesion. How preventive revascularization of noninfarct-related arteries is protective against death is not fully understood; in particular, it is not known whether this beneficial effect could be related to a reduction in heart failure (HF) after MI. Understanding this association is important because HF remains frequent after MI despite widespread use of acute revascularization.4-6 The mechanisms linking acute MI to HF development may theoretically be envisioned as direct sequelae of the MI, including loss of functioning myocytes, development of myocardial fibrosis, and subsequent left ventricular (LV) remodeling adversely affecting ventricular function.7 The overall atherosclerotic burden could also be evoked as a mechanism of post-MI HF. Indeed, chronic myocardial dysfunction resulting from hypoperfusion and/or hibernation may increase the risk of HF,8 particularly if superimposed on a ventricle with irreversibly damaged myocardium.9,10 Although these complex putative mechanisms are challenging to explore clinically, a pragmatic approach is to study the association between atherosclerotic burden and HF in a cohort of patients with acute MI in whom comprehensive follow-up can account for recurrent MI. A clearer clinical appraisal of the determinants of HF after MI could support consideration of revascularization that would extend beyond the acute treatment of the culprit lesion to prevent HF. We therefore evaluated the association between angiographic CAD and subsequent HF in a well-defined community cohort of patients with incident (first-ever) MI. Furthermore, community-based studies11-15 have shown that CAD, diagnosed based on a history of MI, revascularization, or electrocardiographic changes, is common in HF with preserved ejection fraction (HFpEF)—CAD is present in 40% to 50% of patients with HFpEF. To assess the clinical relevance of CAD in the genesis of HF with HFpEF, we examined the prognostic role of CAD according to HF subtype: HF with reduced EF (HFrEF) and HFpEF.

Box Section Ref ID

Key Points

  • Question Does the extent of angiographic coronary artery disease measured at the time of first myocardial infarction influence the occurrence of heart failure?

  • Findings The extent of coronary artery disease is associated with heart failure development after myocardial infarction. The association is independent of intermediate occurrence of recurrent myocardial infarction and does not differ appreciably by reduced or preserved ejection fraction heart failure.

  • Meaning These data shed light on the mechanisms involved in the association between atherosclerotic burden and heart failure risk and underscore the need to further investigate the processes taking place in the transition from myocardial injury to heart failure.

Methods
Study Design and Setting

This population-based study was conducted in Olmsted County, Minnesota (2014 population, approximately 150 287),16 a setting well suited for disease association research owing to its relative isolation from other metropolitan centers and because complete medical records from all sources of care for the local population are indexed and linked via the Rochester Epidemiology Project.17 Because virtually all Olmsted County residents are represented in this system, this data source provides a nearly complete enumeration of the source population for many decades.18 After approval as a minimal-risk study by the Mayo Clinic and Olmsted Medical Center institutional review boards, the study was carried out using the above resource. All persons included in the study provided written authorization for use of their medical records for research.

Cohort Identification and Validation

Residents admitted to Olmsted County hospitals with possible MI from January 1, 1990, to December 31, 2010, were identified with methods previously described.19 Briefly, all events with International Classification of Diseases, Ninth Revision (ICD-9) code 410 (acute MI) were reviewed. In addition, a random sample of events with ICD-9 code 411 (other ischemic heart disease) were reviewed (50% random sample through 1998, 10% random sample from 1999 through 2002, and 100% sample from 2003 through 2010). Additional codes were not included because of their low yield. Myocardial infarctions were validated using standard epidemiologic criteria. Patients with MI diagnosed before 1990 were excluded so that only incident (first-ever) cases were studied. The diagnosis of MI was verified based on the presence of 2 of 3 of the following variables: cardiac pain, electrocardiographic changes, and elevated biomarker levels. Cases were reviewed to ensure that there were no alternative causes for biomarker level elevation.

Primary Exposure Measure

Registries of all coronary angiography procedures—diagnostic and therapeutic—performed in Olmsted County have been maintained since 1979. Because Mayo Clinic is the sole provider of coronary angiography in the county, complete retrieval is possible via the Mayo Clinic Coronary Care Unit database. The primary exposure variable was the extent of CAD as expressed by the number of major coronary arteries with significant obstruction (0-, 1-, 2-, or 3-vessel disease) obtained from coronary angiograms at a median (25th-75th percentile) of 0 (0-1) days after MI. A significant obstruction was defined as angiographic evidence of 50% or more luminal stenosis of any of the epicardial coronary vessels, including side branches.20

Additional Covariates

The medical record was reviewed to determine cardiovascular risk factors, comorbid conditions, MI characteristics, and acute interventions at the time of incident MI. Cigarette use was classified as current, past, or never. Body mass index (BMI) (calculated as weight in kilograms divided by height in meters squared) was determined using the current weight and earliest adult height. Clinical definitions were used to identify hypertension, diabetes mellitus, and hyperlipidemia. Overall comorbidity burden was assessed by the Charlson Comorbidity Index,21 which consists of 17 serious comorbid conditions weighted according to the degree to which they indicate the probability of mortality. The Modification of Diet in Renal Disease equation22 was used to estimate glomerular filtration rate, and ST-segment elevation, anterior MI, and Killip class23 were recorded. Killip class was determined within 24 hours of index MI and was analyzed as a categorical variable (class >1 vs class 1). Acute interventions included reperfusion (thrombolytic therapy or percutaneous coronary intervention) and coronary artery bypass grafting during the index hospitalization.

Outcome Measures

The primary end point was time to incident HF. Participants were followed up using their complete inpatient and outpatient medical records in the community from the index MI (January 1, 1990, to December 31, 2010) to HF incidence, death, or the most recent clinical contact (last follow-up, March 31, 2013). Participants with HF diagnosed by ICD-9 code 428 were identified. Abstractors then reviewed records to validate HF using the Framingham criteria.24 These criteria require the presence of at least 2 of 9 major criteria, or 1 major criterion in addition to 2 of 7 minor criteria, to confirm HF. This approach has been applied previously25 and demonstrated minimal missing data and excellent interobserver agreement. The type of HF was defined according to echocardiographic measurement as HFrEF (EF <50%) and HFpEF (EF ≥50%). Ejection fraction was measured using an approach that was previously described.26 The EF measurement that was closest to the HF diagnosis (applying a predefined maximum period of 60 days) was recorded for each participant. The cutoff of 50% to define preserved and reduced EF was selected according to the guidelines.27 Death (occurrence and date) was ascertained by multiple sources, including autopsy reports, death certificates filed in Olmsted County, obituary notices, and electronic death certificates obtained from the Section of Vital Statistics, Minnesota Department of Health. Recurrent MI data (occurrence and date) were obtained via the Rochester Epidemiology Project on the basis of clinical diagnoses.18

Statistical Analysis

Baseline characteristics across CAD categories are presented as mean (SD) for continuous variables and as frequencies for categorical variables. Cox proportional hazards regression models28 were constructed to estimate the hazard ratios (HRs) and 95% CIs for HF incidence in CAD categories. Several adjustment methods were applied. First, a traditional multivariable adjustment was performed with age, sex, year of index MI, BMI, Charlson Comorbidity Index, smoking status, hypertension, hyperlipidemia, atrial fibrillation, Killip class, ST-segment elevation MI, anterior MI, coronary artery bypass grafting, reperfusion, and estimated glomerular filtration rate as covariates in the model. Second, a propensity score was constructed using multinomial logistic regression through which the probability of being classified into a specific CAD category (0-1, 2, or 3 diseased vessels), conditional on observed baseline covariates, was estimated. Baseline covariates in the latter model included various sociodemographic measures, cardiovascular risk factors, MI characteristics, comorbid conditions, laboratory data, and clinical interventions. Inverse probability weights were calculated using the propensity score29,30 to create a synthetic sample in which the distribution of measured baseline covariates is independent of CAD category, thus accounting for differences between the patients in the CAD categories that could influence the outcome. Because of the instability that can be induced by extreme weights, stabilized weights were used that also preserve the original sample size. Truncation was additionally applied by resetting observations with weights below the 1st percentile and above the 99th percentile to the values of the 1st and 99th percentiles, respectively.29,30 In addition, since standard survival indicators might produce biased estimates in the presence of competing risks,31 the Fine and Gray subdistribution hazard regression model32 was used, with death and recurrent MI treated as competing events. Multivariable-adjusted cumulative incidence rates across CAD categories were based on the abovementioned competing risks model and were estimated using the direct adjustment method.33 Survival analyses were repeated with HFrEF and HFpEF as individual outcomes. The proportional hazards assumption was tested using different approaches and was found not to be violated. Missing values did not exceed 2% in any of the variables other than EF (107 of 588 patients who developed HF [18.2%]). When applicable, an indicator variable reflecting unknown EF was used. Significance was determined using a 1-df Wald χ2 test for trend. Analyses were performed using SAS, version 9.4 (SAS Institute Inc).

Results

Between January 1, 1990, and December 31, 2010, a total of 2943 residents of Olmsted County, Minnesota, were hospitalized with their first MI. Among these, 347 patients (11.8%) had a history of prior HF, 601 (20.4%) had no angiographic assessment available at study entry, and 73 (2.5%) had missing data on important variables and were therefore excluded, leaving 1922 participants in the present study (mean [SD] age, 64 [13] years; 1258 [65.4%] were men).

Classified according to the number of occluded coronary arteries, 692 residents (36.0%) had 3-vessel CAD, 595 (31.0%) had 2-vessel CAD, 566 (29.4%) had 1-vessel CAD, and 69 (3.6%) showed no evidence of substantial coronary occlusion. For analytical purposes, the last 2 groups were combined. Patients with more extensive angiographic CAD were older, presented with greater comorbidities, had a worse cardiovascular risk factor profile, and had a higher Killip class. They were more likely to undergo coronary artery bypass grafting and less likely to undergo reperfusion compared with patients with fewer diseased vessels (Table 1). Stabilized weighting using the inverse propensity score of being classified into a specific CAD category resulted in achieving balance in baseline characteristics between the groups (Table 1).

During a mean (SD) follow-up period of 6.7 (5.9) years, 588 patients (30.6%) developed HF (78 [13.2%] of these in the outpatient setting). Of these, 295 (50.2%) had HFrEF, 186 had HFpEF (31.6%), and 107 (18.2%) had no EF assessment available. The cumulative incidence rates of HF during follow-up across CAD categories, treating death and recurrent MI as competing risks, are depicted in the Figure. The cumulative incidence rates of HF among patients with 0 or 1, 2, and 3 diseased vessels were 10.7%, 14.6%, and 23.0% at 30 days; and 14.7%, 20.6%, and 29.8% at 5 years after MI, respectively (P < .001 for trend). Thus, a higher incidence rate with an increasing number of diseased coronary arteries was evident in both short-term and long-term follow-up after MI.

On a relative scale, worse CAD was associated with an increased risk of subsequent HF (Table 2). A strong, dose-response relationship in the unadjusted model was attenuated, but not eliminated, after the traditional multivariable adjustment and the inverse probability weighting using the propensity score. In general, the results of the 2 adjustment methods were similar. Analyzed by HF type, the associations did not differ substantially between HFrEF and HFpEF for either type of adjustment method (Table 2). Further adjustment for recurrent MI as a time-dependent covariate in the multivariable Cox proportional hazards regression model yielded HRs (95% CIs) for HF of 1.24 (0.98-1.57) and 1.66 (1.33-2.09) in patients with 2 and 3 vs 0 or 1 occluded vessels, respectively (P < .001 for trend); the increasing trend in HRs was observed similarly for HFrEF (1.19 [0.85-1.65] and 1.71 [1.24-2.35]; P = .001 for trend) and HFpEF (1.23 [0.81-1.89] and 1.67 [1.11-2.50]; P = .01 for trend), respectively. Thus, although recurrent MI was associated with HF risk (HR, 4.01; 95% CI, 3.03-5.30), it did not substantially confound the association between CAD and HF. Similarly, infarct size, as estimated by serum creatine kinase MB and cardiac troponin T values, has been shown34 to be associated with HF risk after MI, but no association was observed in the present study between the biomarkers and more extensive CAD (P = .64 and P = .91 for trend for serum creatine kinase MB and cardiac troponin T values, respectively). Thus, infarct size also did not substantially confound the association between CAD and HF. Effect modifications of year of index MI, age, and sex on the association between CAD extent and HF were assessed and rejected (all P > .10).

In ancillary analyses, results from the Fine and Gray32 subdistribution hazard regression models, with death and recurrent MI treated as competing events, were similar to those obtained from the multivariable-adjusted models. Reported as HR (95% CI), once again, a dose-response association was observed between CAD extent and HF in patients with 2 and 3 compared with 0 or 1 occluded vessels, respectively (1.24 [0.96-1.59] and 1.55 [1.21-1.99]; P < .001 for trend). When analyzed by type of HF, similar results were seen for HFrEF (1.26 [0.88-1.81] and 1.66 [1.17-2.34]; P = .004 for trend), and a greater attenuation was observed for HFpEF (1.06 [0.68-1.65] and 1.28 [0.83-1.98]; P = .24 for trend).

In a second ancillary analysis, patients with no evidence of coronary artery occlusion (n = 69) were excluded from the analysis. The results were almost identical to those presented herein.

Discussion

The present study reports the comprehensive experience of an entire community, with nearly 2000 patients with validated first MI, no prior HF, and baseline angiographic assessment who were followed up longitudinally for a mean duration of 6.7 years. During this period, 30.6% of the participants developed incident HF. Patients with greater atherosclerotic burden at baseline tended to be older and have a higher Killip class and worse cardiovascular profiles; their HF-free survival was substantially shorter than that of patients with lower atherosclerotic burden in a clear dose-response fashion. With the use of comprehensive statistical adjustment methods to balance the baseline characteristics of the compared groups and accounting for competing risks during follow-up, the association between angiographic CAD extent and HF incidence remained robust, statistically significant, and clinically substantial. When evaluated by HF type, the magnitude of the association did not differ substantially according to preserved or reduced EF. Thus, an increasing extent of CAD, as detected by angiography at the time of the first MI, is an indicator of HF incidence during long-term follow-up. The association manifests itself promptly after MI, is independent of recurrent MI, and applies similarly to HFrEF and HFpEF.

The association between angiographically determined CAD extent and post-MI HF was previously reported in 2 studies.35,36 Among 1619 patients enrolled in the Thrombolysis and Angioplasty in Myocardial Infarction trials,35 the number of diseased vessels was indicative of HF development, both in the hospital and during 1-year follow-up. In the Valsartan in Acute Myocardial Infarction Study36 among patients with acute MI complicated by either clinical or radiologic signs of HF, Janardhanan et al suggested a prognostic role for increasing extent of CAD in adverse cardiovascular outcomes, including HF exacerbation. Although these findings are important, caution should be used in interpreting them. Because both studies examined selected populations of clinical trial participants, the generalizability of their results is uncertain.37 Moreover, the results of these 2 studies now reflect somewhat dated cohorts that do not capture major changes in the epidemiology of MI that occurred in the past 2 decades, including the increased proportion of non–ST-segment elevation MI, improved treatment, reduced short-term case fatality and recurrent MI rates, and increased proportion of morbidity and mortality due to noncardiovascular causes.19,38,39 Temporal trends have also occurred in HF complicating MI, with a decline in its incidence4-6 and a change in the case mix with an increasing proportion of HFpEF.4

The clinical implications of CAD among 376 patients with HFpEF were recently reported.11 Angiographically proved CAD, which was present in 255 (67.8%) of the participants, was associated with increased mortality and greater deterioration in ventricular function. A worse prognosis associated with an increasing extent of CAD was also noted in HFrEF.40 Although these studies demonstrated the prognostic role of angiographic CAD in symptomatic patients with HFrEF and HFpEF, no other study to date, to our knowledge, has evaluated the association between angiographic CAD at the time of acute MI and subsequent risk for HF according to EF. In the present study, we addressed this gap in knowledge and provided evidence that the number of diseased vessels, as defined angiographically at the time of the first-ever MI, is a strong indicator of both HFrEF and HFpEF.

The mechanisms through which concomitant atherosclerosis in coronary vessels other than the culprit artery adversely affect HF risk post-MI need further study. Diffuse atherosclerosis may directly or indirectly exert an adverse effect on long-term prognosis of MI, either because of the extent of ischemic damage or by causing subsequent events (eg, recurrent MI) that increase the risk of HF. Because we captured recurrent MIs during follow-up in our community cohort and accounted for recurrent MIs analytically, our results do not give credence to recurrent MI being a determinant of the CAD-HF association. Left ventricular dilation has been shown41 to occur frequently after primary percutaneous coronary intervention in patients with acute MI despite sustained patency of the infarct-related artery and preservation of regional and global LV function. This finding, although highlighting the importance of optimal microvascular flow and tissue reperfusion, also suggests that factors different from infarct size and culprit vessel patency may play a role in post-MI LV remodeling and subsequent HF. Patients with both epicardial and endocardial CAD may have chronic hypoperfusion that leads to increased myocardial stiffness secondary to chronic inflammation and fibrosis. The increased myocardial stiffness, in turn, may impair systolic and diastolic function.10 These findings resonate with reports36 of an association between CAD and HFpEF.

Some limitations of our study should be acknowledged to aid in data interpretation. As in any observational study, we cannot rule out the effect of residual confounding due to unmeasured variables. Of the potentially eligible patients, 601 (20.4%) did not have an angiographic assessment during the index hospitalization, which may limit the generalizability of this sample. Of the participants with HF, 106 (17.9%) were missing EF data. These results emanate from a single community of mostly white race/ethnicity, which may limit the generalizability to groups not adequately represented.

The present investigation has several notable strengths. We capitalized on the comprehensive data resources of the Rochester Epidemiology Project to examine the role of angiographically determined CAD on post-MI prognosis in the community. We report on a large, population-based inception cohort registered at the time of their first MI validated by standardized criteria.19 Heart failure, the primary outcome measure, was rigorously ascertained and validated using established criteria.24 Echocardiographic data allowed categorization as HFrEF and HFpEF, which is important to understand the HF syndrome. Finally, different analytical methods were used to estimate the net effect of CAD on HF risk, all yielding similar results, which attest to the robustness of our findings.

Conclusions

During the past 2 decades, major changes in the epidemiology of MI have occurred. Progress in its acute treatment improved short-term survival, but HF remains frequent after MI and leads to excess mortality. Hence, the acute treatment of MI aimed at restoring vessel patency is not sufficient to prevent HF, underscoring the importance of understanding the contemporary mechanisms leading to its development. Our study provides insight into the prognostic role of the extent of CAD at the time of first MI in the development of HF as well as on the mechanisms involved. The present findings underscore the importance of further investigations into processes taking place in the transition from the initial myocardial injury to HF. Understanding this transition is crucial to prevent the development of HF after MI.

Back to top
Article Information

Corresponding Author: Véronique L. Roger, MD, MPH, Division of Cardiovascular Diseases, Department of Health Sciences Research, Mayo Clinic, 200 First St SW, Rochester, MN 55905 (roger.veronique@mayo.edu).

Accepted for Publication: January 15, 2016.

Published Online: March 30, 2016. doi:10.1001/jamacardio.2016.0074.

Author Contributions: Dr Roger had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Gerber, Enriquez-Sarano, Chamberlain, Roger.

Acquisition, analysis, or interpretation of data: Gerber, Weston, Enriquez-Sarano, Manemann, Jiang, Roger.

Drafting of the manuscript: Gerber, Manemann, Roger.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Gerber, Weston, Jiang, Roger.

Obtained funding: Roger.

Administrative, technical, or material support: Roger.

Study supervision: Gerber, Enriquez-Sarano, Roger.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Enriquez-Sarano reports grants from Edwards LLC, outside the submitted work. Dr Roger reports grants from National Institutes of Health and grants from National Institutes on Aging during the conduct of this study. No other disclosures were reported.

Funding/Support: This study was supported by National Institutes of Health grants R01 HL59205, R01 HL72435, and R01 HL120957, as well as by the Rochester Epidemiology Project and grant R01 AG034676 from the National Institute on Aging.

Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

References
1.
Wald  DS, Morris  JK, Wald  NJ,  et al; PRAMI Investigators.  Randomized trial of preventive angioplasty in myocardial infarction.  N Engl J Med. 2013;369(12):1115-1123.PubMedGoogle ScholarCrossref
2.
Mauri  L.  Nonculprit lesions—or guilty by association.  N Engl J Med. 2013;369(12):1166-1167.PubMedGoogle ScholarCrossref
3.
El-Hayek  GE, Gershlick  AH, Hong  MK,  et al.  Meta-analysis of randomized controlled trials comparing multivessel versus culprit-only revascularization for patients with ST-segment elevation myocardial infarction and multivessel disease undergoing primary percutaneous coronary intervention.  Am J Cardiol. 2015;115(11):1481-1486.PubMedGoogle ScholarCrossref
4.
Gerber  Y, Weston  SA, Berardi  C,  et al.  Contemporary trends in heart failure with reduced and preserved ejection fraction after myocardial infarction: a community study.  Am J Epidemiol. 2013;178(8):1272-1280.PubMedGoogle ScholarCrossref
5.
McManus  DD, Chinali  M, Saczynski  JS,  et al.  30-Year trends in heart failure in patients hospitalized with acute myocardial infarction.  Am J Cardiol. 2011;107(3):353-359.PubMedGoogle ScholarCrossref
6.
Chen  J, Hsieh  AF, Dharmarajan  K, Masoudi  FA, Krumholz  HM.  National trends in heart failure hospitalization after acute myocardial infarction for Medicare beneficiaries: 1998-2010.  Circulation. 2013;128(24):2577-2584.PubMedGoogle ScholarCrossref
7.
Sutton  MG, Sharpe  N.  Left ventricular remodeling after myocardial infarction: pathophysiology and therapy.  Circulation. 2000;101(25):2981-2988.PubMedGoogle ScholarCrossref
8.
Wijns  W, Vatner  SF, Camici  PG.  Hibernating myocardium.  N Engl J Med. 1998;339(3):173-181.PubMedGoogle ScholarCrossref
9.
Gheorghiade  M, Bonow  RO.  Chronic heart failure in the United States: a manifestation of coronary artery disease.  Circulation. 1998;97(3):282-289.PubMedGoogle ScholarCrossref
10.
Velagaleti  RS, Vasan  RS.  Heart failure in the twenty-first century: is it a coronary artery disease or hypertension problem?  Cardiol Clin. 2007;25(4):487-495.PubMedGoogle ScholarCrossref
11.
Hwang  SJ, Melenovsky  V, Borlaug  BA.  Implications of coronary artery disease in heart failure with preserved ejection fraction.  J Am Coll Cardiol. 2014;63(25, pt A):2817-2827.PubMedGoogle ScholarCrossref
12.
Owan  TE, Hodge  DO, Herges  RM, Jacobsen  SJ, Roger  VL, Redfield  MM.  Trends in prevalence and outcome of heart failure with preserved ejection fraction.  N Engl J Med. 2006;355(3):251-259.PubMedGoogle ScholarCrossref
13.
Lam  CS, Donal  E, Kraigher-Krainer  E, Vasan  RS.  Epidemiology and clinical course of heart failure with preserved ejection fraction.  Eur J Heart Fail. 2011;13(1):18-28.PubMedGoogle ScholarCrossref
14.
Bhatia  RS, Tu  JV, Lee  DS,  et al.  Outcome of heart failure with preserved ejection fraction in a population-based study.  N Engl J Med. 2006;355(3):260-269.PubMedGoogle ScholarCrossref
15.
Tribouilloy  C, Rusinaru  D, Mahjoub  H,  et al.  Prognosis of heart failure with preserved ejection fraction: a 5 year prospective population-based study.  Eur Heart J. 2008;29(3):339-347.PubMedGoogle ScholarCrossref
16.
United States Census. State & County QuickFacts. http://quickfacts.census.gov/qfd/states/27/27109.html. Accessed October 23, 2015.
17.
Melton  LJ  III.  History of the Rochester Epidemiology Project.  Mayo Clin Proc. 1996;71(3):266-274.PubMedGoogle ScholarCrossref
18.
St Sauver  JL, Grossardt  BR, Yawn  BP, Melton  LJ  III, Rocca  WA.  Use of a medical records linkage system to enumerate a dynamic population over time: the Rochester Epidemiology Project.  Am J Epidemiol. 2011;173(9):1059-1068.PubMedGoogle ScholarCrossref
19.
Roger  VL, Weston  SA, Gerber  Y,  et al.  Trends in incidence, severity, and outcome of hospitalized myocardial infarction.  Circulation. 2010;121(7):863-869.PubMedGoogle ScholarCrossref
20.
Scanlon  PJ, Faxon  DP, Audet  AM,  et al.  ACC/AHA guidelines for coronary angiography: a report of the American College of Cardiology/American Heart Association Task Force on practice guidelines (Committee on Coronary Angiography): developed in collaboration with the Society for Cardiac Angiography and Interventions.  J Am Coll Cardiol. 1999;33(6):1756-1824.PubMedGoogle ScholarCrossref
21.
Charlson  ME, Pompei  P, Ales  KL, MacKenzie  CR.  A new method of classifying prognostic comorbidity in longitudinal studies: development and validation.  J Chronic Dis. 1987;40(5):373-383.PubMedGoogle ScholarCrossref
22.
Levey  AS, Coresh  J, Greene  T,  et al; Chronic Kidney Disease Epidemiology Collaboration.  Using standardized serum creatinine values in the Modification of Diet in Renal Disease study equation for estimating glomerular filtration rate.  Ann Intern Med. 2006;145(4):247-254.PubMedGoogle ScholarCrossref
23.
Killip  T  III, Kimball  JT.  Treatment of myocardial infarction in a coronary care unit. A two year experience with 250 patients.  Am J Cardiol. 1967;20(4):457-464.PubMedGoogle ScholarCrossref
24.
McKee  PA, Castelli  WP, McNamara  PM, Kannel  WB.  The natural history of congestive heart failure: the Framingham study.  N Engl J Med. 1971;285(26):1441-1446.PubMedGoogle ScholarCrossref
25.
Roger  VL, Weston  SA, Redfield  MM,  et al.  Trends in heart failure incidence and survival in a community-based population.  JAMA. 2004;292(3):344-350.PubMedGoogle ScholarCrossref
26.
Dunlay  SM, Roger  VL, Weston  SA, Jiang  R, Redfield  MM.  Longitudinal changes in ejection fraction in heart failure patients with preserved and reduced ejection fraction.  Circ Heart Fail. 2012;5(6):720-726.PubMedGoogle ScholarCrossref
27.
Yancy  CW, Jessup  M, Bozkurt  B,  et al; Writing Committee Members; American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines.  2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association Task Force on practice guidelines.  Circulation. 2013;128(16):e240-e327.PubMedGoogle ScholarCrossref
28.
Cox  DR.  Regression analysis and life table.  J R Stat Soc B. 1972;34:187-222.Google Scholar
29.
Austin  PC.  The use of propensity score methods with survival or time-to-event outcomes: reporting measures of effect similar to those used in randomized experiments.  Stat Med. 2014;33(7):1242-1258.PubMedGoogle ScholarCrossref
30.
Hernán  MA, Robins  JM.  Causal Inference. Boca Raton, FL: Chapman & Hall/CRC; 2016, forthcoming.
31.
Wolbers  M, Koller  MT, Witteman  JC, Steyerberg  EW.  Prognostic models with competing risks: methods and application to coronary risk prediction.  Epidemiology. 2009;20(4):555-561.PubMedGoogle ScholarCrossref
32.
Fine  JP, Gray  RJ.  A proportional hazards model for the subdistribution of a competing risk.  J Am Stat Assoc. 1999;94(446):496-509.Google ScholarCrossref
33.
Zhang  X, Zhang  MJ.  SAS macros for estimation of direct adjusted cumulative incidence curves under proportional subdistribution hazards models.  Comput Methods Programs Biomed. 2011;101(1):87-93.PubMedGoogle ScholarCrossref
34.
Gerber  Y, Jaffe  AS, Weston  SA, Jiang  R, Roger  VL.  Prognostic value of cardiac troponin T after myocardial infarction: a contemporary community experience.  Mayo Clin Proc. 2012;87(3):247-254.PubMedGoogle ScholarCrossref
35.
O’Connor  CM, Hathaway  WR, Bates  ER,  et al.  Clinical characteristics and long-term outcome of patients in whom congestive heart failure develops after thrombolytic therapy for acute myocardial infarction: development of a predictive model.  Am Heart J. 1997;133(6):663-673.PubMedGoogle ScholarCrossref
36.
Janardhanan  R, Kenchaiah  S, Velazquez  EJ,  et al; VALIANT Investigators.  Extent of coronary artery disease as a predictor of outcomes in acute myocardial infarction complicated by heart failure, left ventricular dysfunction, or both.  Am Heart J. 2006;152(1):183-189.PubMedGoogle ScholarCrossref
37.
Steg  PG, López-Sendón  J, Lopez de Sa  E,  et al; GRACE Investigators.  External validity of clinical trials in acute myocardial infarction.  Arch Intern Med. 2007;167(1):68-73.PubMedGoogle ScholarCrossref
38.
Yeh  RW, Sidney  S, Chandra  M, Sorel  M, Selby  JV, Go  AS.  Population trends in the incidence and outcomes of acute myocardial infarction.  N Engl J Med. 2010;362(23):2155-2165.PubMedGoogle ScholarCrossref
39.
Gerber  Y, Weston  SA, Jiang  R, Roger  VL.  The changing epidemiology of myocardial infarction in Olmsted County, Minnesota, 1995-2012.  Am J Med. 2015;128(2):144-151.PubMedGoogle ScholarCrossref
40.
Felker  GM, Shaw  LK, O’Connor  CM.  A standardized definition of ischemic cardiomyopathy for use in clinical research.  J Am Coll Cardiol. 2002;39(2):210-218.PubMedGoogle ScholarCrossref
41.
Bolognese  L, Neskovic  AN, Parodi  G,  et al.  Left ventricular remodeling after primary coronary angioplasty: patterns of left ventricular dilation and long-term prognostic implications.  Circulation. 2002;106(18):2351-2357.PubMedGoogle ScholarCrossref
×