1. INTRODUCTION

According to the World Health Organization, ischemic heart disease has been the leading cause of death worldwide in recent years. The mortality rate is 8.93 million deaths annually, accounting for nearly 16% of all-cause mortality.1 Despite the standard medical treatment recommended by American Heart Association/European Society of Cardiology, and the evolution of interventional therapy, acute myocardial infarction (AMI) still causes tremendous morbidity and mortality, thereby burdening the healthcare system.2 Thus, the factors related to AMI influencing mortality should be addressed.

Chronic obstructive pulmonary disease (COPD), with a prevalence rate of approximately 11.7%, causes serious and long-term disability and early death, accounting for approximately 3.2 million deaths annually.1 COPD and AMI share common risk factors that promote inflammation, characterized by the enhanced production of chemokines and cytokines such as interleukin (IL)-1, IL-6, and tumor necrosis factor-alpha, which enter the circulation from the lungs. Moreover, inflammation deteriorates preexisting comorbidities and may initiate lung cancer.3

Despite the pharmacological management of AMI, systemic inflammation due to COPD promotes atherosclerosis, endothelial dysfunction, and arterial stiffness;4 thereby reducing treatment efficacy and survival in patients with AMI. Meanwhile, the main underlying causes of COPD in Asian populations are a high prevalence rate of cigarette smoking and a very high level of air pollution.5 Whether the different causes of COPD influence the outcome of AMI needs clarification.

Furthermore, the medications for secondary prevention after AMI, such as beta-blockers, have been associated with reduced mortality in patients with COPD.6 On the other hand, the safety of COPD medications, such as inhaled bronchodilators and corticosteroids, has yet to be investigated in patients with AMI.7,8 Therefore, this nationwide population-based cohort study aimed to investigate the long-term impact of COPD and the safety of COPD medications in patients with AMI.

2. METHODS 2.1. Data collection

We conducted a nationwide cohort study in Taiwan, including all hospitalized patients with a primary diagnosis of AMI, which was approved by the Human Research Committee of Kaohsiung Veterans General Hospital. Data were retrieved from a full-population dataset from Taiwan’s National Health Insurance Research Database (NHIRD). Taiwan launched the National Health Insurance (NHI) program in 1995 and enrolled 99.9% of citizens and legal residents. This dataset contains data for approximately 23 million people whose registration files, diagnosis codes, medications, examinations, and procedures were recorded for reimbursement and research.9 The accuracy of claimed diagnosis codes have been validated in different studies.10

2.2. Study design

We identified all hospitalized patients with a primary diagnosis of AMI (International Classification of Diseases, Ninth Revision [ICD-9], 410-410.92) for the first time between January 2000 and December 2012 in Taiwan. Patients aged <18 years, >120 years, and those with unclear insurance records were excluded. In addition, patients with AMI and a concurrent diagnosis of COPD (ICD-9 codes 491, 492, or 496) were included. Propensity score matching was performed using a 1:1 matching protocol based on sex, age group, comorbidities, and interventions (percutaneous coronary intervention, coronary artery bypass graft, and intra-aortic balloon pump). Patients with AMI were further divided into ST-elevation myocardial infarction (STEMI) and non-STEMI (NSTEMI) subtypes. Patients with AMI diagnosed as both STEMI and NSTEMI were excluded. Patients with either subtype in the AMI cohort with a concurrent diagnosis of COPD were identified individually by ICD-9 codes, whereas patients with either subtype in the AMI cohort without COPD were selected by propensity score matching and designated as the control group. We used ICD-9 codes to identify AMI subtypes and comorbidities and anatomical therapeutic chemical codes to analyze the survival influence of COPD and AMI medications during hospitalization.

2.3. Outcome analysis

All enrolled patients were followed up until December 31, 2012, or death, whichever occurred first. The study’s primary endpoint was mortality, defined as the end date of NHI coverage. The criteria for mortality are reliable in Taiwan because NHI is mandatorily paid monthly, even in low-income households (the government offers premium subsidies). Thus, the difference between the date of admission and the end date of NHI coverage should be within one month.

2.4. Statistical analyses

Data were analyzed using SAS software (version 9.4; SAS Institute, Inc., Cary, NC). Categorical data are presented as percentages and compared using Chi-square tests. Continuous variables are presented as means and SDs and compared using paired t-tests. A Cox proportional hazards model was used to calculate hazard ratios and 95% CIs. Kaplan-Meier analysis was performed to estimate the cumulative survival and differences between patients with AMI and COPD and the control group. A log-rank test was performed to evaluate differences between curves. Differences with a two-tailed p value <0.05 were considered statistically significant.

3. RESULTS 3.1. Descriptive characteristics of the study groups

Between January 2000 and December 2012, 186 326 patients with a primary diagnosis of AMI for the first time were hospitalized. Based on the exclusion criteria, 214 patients were excluded (Fig. 1), and 186 112 patients with AMI were included. Patients with AMI were categorized into STEMI (73 148 patients) and NSTEMI (112 408 patients) cohorts for further analysis. Patients with a concomitant COPD diagnosis were selected. There were 23 704 patients in the overall AMI cohort (6569 in the STEMI cohort and 17 089 in the NSTEMI cohort). Patients without COPD were propensity score matched (1:1) for each cohort (Fig. 1). Patient demographic and clinical characteristics (Table 1) revealed that, in both groups of all cohorts, most patients were male and >65 years of age. No significant differences were observed in comorbidities and coronary vessel interventions between groups, except in the AMI cohort, which had a higher proportion of patients with STEMI and concomitant COPD (32.29% vs. 27.7%; p < 0.0001). In each cohort, patients with COPD were administered more calcium channel blockers, short- and long-acting bronchodilators, Xanthiums, and corticosteroids than those without COPD. In addition, patients without COPD received more antiplatelets, angiotensin-converting enzyme inhibitors (ACEIs), angiotensin receptor blockers (ARB), statins, beta-blockers, and heparin or low-molecular-weight heparin than patients with COPD. No significant differences were observed in the use of vasopressors (dopamine or norepinephrine) and interventions (percutaneous coronary intervention, coronary artery bypass graft, and intra-aortic balloon pump) between the two groups.

Table 1 - Demographic and clinical characteristics of AMI, STEMI, and NSTEMI patients with and without COPD Variables All AMI p STEMI p NSTEMI p With COPD Without COPD With COPD Without COPD With COPD Without COPD (n = 23 704) (n = 23 704) (n = 6569) (n = 6569) (n = 17 089) (n = 17 089) Men, n (%) 17524 (73.93) 17514 (73.89) 0.9167 5149 (78.38) 5144 (78.31) 0.9157 12342 (72.22) 12336 (72.19) 0.9422 Age ≥65 y, n (%) 21033 (88.73) 21041 (88.77) 0.9074 5694 (86.68) 5698 (86.74) 0.9181 15299 (89.53) 15308 (89.58) 0.8735 Comorbidities, n (%)  Hypertension 16810 (70.92) 16813 (70.93) 0.9758 4214 (64.15) 4212 (64.12) 0.971 12564 (73.52) 12559 (73.49) 0.9511  Dyslipidemia 7720 (32.57) 7707 (32.51) 0.8986 2164 (32.94) 2154 (32.79) 0.8527 5542 (32.43) 5522 (32.31) 0.8172  Diabetes mellitus 9334 (39.38) 9373 (39.54) 0.7140 2327 (35.42) 2349 (35.76) 0.6885 6992 (40.92) 7026 (41.11) 0.7085  Peripheral vascular disease 1294 (5.46) 1276 (5.38) 0.7150 272 (4.14) 270 (4.11) 0.9301 1016 (5.95) 1011 (5.92) 0.9088  Heart failure 3783 (15.96) 3782 (15.96) 0.9900 722 (10.99) 715 (10.88) 0.8449 3057 (17.89) 3037 (17.77) 0.7775  End-stage renal disease 839 (3.54) 840 (3.54) 0.9802 145 (2.21) 144 (2.19) 0.9526 693 (4.06) 692 (4.05) 0.9781  Cerebrovascular accident 1000 (4.22) 998 (4.21) 0.9635 253 (3.85) 246 (3.74) 0.7494 746 (4.37) 739 (4.32) 0.8527 STEMI 6568 (27.71) 7654 (32.29) <0.0001 Interventions, n (%)  Percutaneous coronary intervention 7422 (31.31) 7425 (31.32) 0.9763 3130 (47.65) 3123 (47.54) 0.9027 4274 (25.01) 4263 (24.95) 0.8907  Coronary artery bypass graft 1261 (5.32) 1263 (5.33) 0.9674 389 (5.92) 390 (5.94) 0.9705 866 (5.07) 867 (5.07) 0.9803  Intra-aortic balloon pump 1021 (4.31) 1019 (4.30) 0.9639 414 (6.30) 412 (6.27) 0.9427 604 (3.53) 607 (3.55) 0.9301 Medications, n (%)  Antiplatelet 18437 (77.78) 19202 (81.01) <0.0001 5609 (85.39) 5789 (88.13) <0.0001 12790 (74.84) 13375 (78.27) <0.0001  ACEI or ARB 12463 (52.58) 13331 (56.24) <0.0001 4112 (62.60) 4248 (64.67) 0.0136 8323 (48.70) 9006 (52.70) <0.0001  Statin 4830 (20.38) 5586 (23.57) <0.0001 1479 (22.51) 1650 (25.12) 0.0005 3340 (19.54) 3940 (23.06) <0.0001  Beta-blocker 7629 (32.18) 10817 (45.63) <0.0001 2455 (37.37) 3335 (50.77) <0.0001 5159 (30.19) 7447 (43.58) <0.0001  Calcium channel blocker 9612 (40.55) 8176 (34.49) <0.0001 2320 (35.32) 1820 (27.71) <0.0001 7280 (42.60) 6507 (38.08) <0.0001  Heparin or LMWH 14210 (59.95) 15408 (65.00) <0.0001 4711 (71.72) 4983 (75.86) <0.0001 9466 (55.39) 10348 (60.55) <0.0001  Dopamine 5070 (21.39) 4998 (21.09) 0.4188 1565 (23.82) 1470 (22.38) 0.0492 3495 (20.45) 3498 (20.47) 0.9679  Norepinephrine 2877 (12.14) 2914 (12.29) 0.2207 657 (10.00) 662 (10.08) 0.8846 2165 (12.67) 2324 (13.60) 0.0109  Spironolactone 3374 (14.23) 2994 (12.63) <0.0001 840 (12.79) 707 (10.76) 0.0003 2529 (14.80) 2311 (13.52) 0.0007  Long-acting muscarinic antagonist 31 (0.47) 4 (0.06) <0.0001 31 (0.47) 4 (0.06) <0.0001 189 (1.11) 12 (0.07) <0.0001  Long-acting beta-agonist 221 (3.36) 32 (0.49) <0.0001 221 (3.36) 32 (0.49) <0.0001 886 (5.18) 128 (0.75) <0.0001  Xanthium 2371 (36.09) 642 (9.77) <0.0001 2371 (36.09) 642 (9.77) <0.0001 6172 (36.12) 1925 (11.26) <0.0001  Corticosteroid 1504 (22.90) 566 (8.62) <0.0001 1504 (22.90) 566 (8.62) <0.0001 5884 (34.43) 2890 (16.91) <0.0001  Short-acting muscarinic antagonist 3050 (46.43) 1386 (21.10) <0.0001 3050 (46.43) 1386 (21.10) <0.0001 9380 (54.89) 5114 (29.93) <0.0001  Short-acting beta-agonist 2384 (36.29) 1067 (16.24) <0.0001 2384 (36.29) 1067 (16.24) <0.0001 6525 (38.18) 3527 (20.64) <0.0001

ACEI = angiotensin-converting enzyme inhibitor; AMI = acute myocardial infarction; ARB = angiotensin receptor blocker; COPD = chronic obstructive pulmonary disease; LMWH = low-molecular-weight heparin; NSTEMI = non-ST-elevated myocardial infarction; STEMI = ST-elevated myocardial infarction.

Fig. 1:

A, Flow chart for the selection of patients with AMI with and without COPD from the Taiwan National Health Insurance Research Database. B, Flow chart for the selection of STEMI and NSTEMI cohorts with and without COPD from the Taiwan National Health Insurance Research Database. AMI = acute myocardial infarction; COPD = chronic obstructive pulmonary disease; NSTEMI = non-ST-elevation myocardial infarction; STEMI = ST-elevated myocardial infarction.

3.2. Survival analysis

During the 12-year follow-up, Kaplan-Meier curves suggested that patients with COPD had a higher mortality risk than those without COPD in all cohorts (AMI, STEMI, and NSTEMI). Differences in mortality remained among subgroups for sex, age, hypertension, diabetes mellitus, and percutaneous coronary intervention (log-rank, all p < 0.0001; Figs. 2–4). The mortality rates (p < 0.0001) were 77.65% and 67.89% in patients with and without COPD in the AMI cohort, respectively; 74.12% and 61.85% in patients with and without COPD in the STEMI cohort, respectively; and 78.99% and 70.52% in patients with and without COPD in the NSTEMI cohort, respectively.

Fig. 2:

Kaplan-Meier survival curve for AMI comparing patients with and without COPD and in different subgroups, including sex, age, hypertension, diabetes mellitus, and percutaneous coronary intervention. AMI = acute myocardial infarction; COPD = chronic obstructive pulmonary disease.

Fig. 3:

Kaplan-Meier survival curve for STEMI comparing patients with and without COPD and in different subgroups including sex, age, hypertension, diabetes mellitus, and percutaneous coronary intervention. COPD = chronic obstructive pulmonary disease; STEMI = ST-elevation myocardial infarction.

Fig. 4:

Kaplan-Meier survival curve for NSTEMI comparing patients with and without COPD and in different subgroups, including sex, age, hypertension, diabetes mellitus, and percutaneous coronary intervention. COPD = chronic obstructive pulmonary disease; NSTEMI = non-ST-elevation myocardial infarction.

Cox proportional hazards regression analysis revealed that the mortality rate in AMI patients was higher in males, those ≥65 years of age, and those with hypertension, diabetes mellitus, peripheral vascular disease, heart failure, end-stage renal disease, and cerebral vascular accidents (Table 2). COPD in patients with AMI was associated with a 12% higher mortality rate (20% higher in patients with STEMI and 7% higher in patients with NSTEMI). The negative impacts of age and comorbidities on mortality risk were similar in all cohorts (Table 2). Post-AMI medications such as antiplatelets, beta-blockers, ACEI/ARB, and statins substantially reduced mortality risk in all cohorts. However, corticosteroids, short-acting beta-agonists (SABA), and short-acting muscarinic antagonists (SAMA) significantly increased mortality risk in the AMI cohort. The adverse effects of corticosteroids and short-acting bronchodilators were consistent in all cohorts. In contrast, long-acting beta-agonists (LABAs), long-acting muscarinic antagonists (LAMA), and Xanthiums significantly increased survival rates in the AMI cohort. Similar results were observed in the NSTEMI cohort; however, no significant survival benefit was observed in the STEMI cohort.

Table 2 - Cox proportional hazards regression analysis in patients with AMI with and without COPD Variables All AMI (n = 23 704) STEMI (n = 6569) NSTEMI(n = 17 089) Adjusted HR (95% CI) p Adjusted HR (95% CI) p Adjusted HR (95% CI) p Male 1.09 (1.06-1.12) <0.0001 0.97 (0.92-1.01) 0.1587 1.10 (1.07-1.13) <0.0001 Age≥65y 2.24 (2.15-2.34) <0.0001 2.45 (2.25-2.66) <0.0001 2.10 (2.00-2.20) <0.0001 Comorbidities  COPD 1.12 (1.09-1.14) <0.0001 1.20 (1.15-1.26) <0.0001 1.07 (1.04-1.10) <0.0001  Hypertension 1.09 (1.07-1.12) <0.0001 1.15 (1.10-1.20) <0.0001 1.08 (1.05-1.11) <0.0001  Diabetes mellitus 1.25 (1.22-1.28) <0.0001 1.31 (1.25-1.37) <0.0001 1.22 (1.18-1.25) <0.0001  Peripheral vascular disease 1.36 (1.30-1.42) <0.0001 1.37 (1.24-1.51) <0.0001 1.32 (1.25-1.38) <0.0001  Heart failure 1.27 (1.23-1.30) <0.0001 1.37 (1.29-1.46) <0.0001 1.26 (1.22-1.30) <0.0001  End-stage renal disease 1.61 (1.52-1.70) <0.0001 1.60 (1.40-1.81) <0.0001 1.54 (1.45-1.64) <0.0001  Cerebral vascular accident 1.06 (1.01-1.12) 0.0227 1.16 (1.05-1.29) 0.0049 1.08 (1.02-1.14) 0.0129 STEMI 0.91 (0.89-0.94) <0.0001 Interventions  Percutaneous coronary intervention 0.50 (0.49-0.51) <0.0001 0.49 (0.47-0.52) <0.0001 0.50 (0.49-0.52) <0.0001  Coronary artery bypass graft 0.51 (0.49-0.54) <0.0001 0.53 (0.48-0.58) <0.0001 0.50 (0.47-0.57) <0.0001  Intra-aortic balloon pump 1.99 (1.88-2.10) <0.0001 2.04 (1.87-2.23) <0.0001 1.85 (1.72-1.98) <0.0001 Medications  Antiplatelet 0.75 (0.73-0.77) <0.0001 0.89 (0.84-0.95) 0.0003 0.71 (0.68-0.73) <0.0001  ACEI or ARB 0.76 (0.47-0.78) <0.0001 0.77 (0.74-0.81) <0.0001 0.74 (0.72-0.77) <0.0001  Statin 0.85 (0.81-0.88) <0.0001 0.83 (0.76-0.90) <0.0001 0.85 (0.81-0.90) <0.0001  Beta-blocker 0.86 (0.84-0.88) <0.0001 0.90 (0.86-0.94) <0.0001 0.85 (0.83-0.88) <0.0001  Long-acting muscarinic antagonist 0.82 (0.69-0.96) 0.0162 0.75 (0.46-1.21) 0.2342 0.80 (0.68-0.96) 0.0143  Long-acting beta-agonist 0.87 (0.81-0.94) 0.0001 0.91 (0.77-1.07) 0.2458 0.89 (0.83-0.97) 0.0044  Xanthium 0.94 (0.91-0.96) <0.0001 0.94 (0.89-1.00) 0.0336 0.93 (0.90-0.96) <0.0001  Corticosteroid 1.10 (1.07-1.14) <0.0001 1.09 (1.02-1.16) 0.0104 1.12 (1.09-1.16) <0.0001  Short-acting antimuscarinic agent 1.30 (1.26-1.34) <0.0001 1.42 (1.33-1.52) <0.0001 1.25 (1.20-1.30) <0.0001  Short-acting beta-agonist 1.20 (1.16-1.23) <0.0001 1.17 (1.09-1.24) <0.0001 1.20 (1.16-1.25) <0.0001

ACEI = angiotensin-converting enzyme inhibitor; AMI = acute myocardial infarction; ARB = angiotensin receptor blocker; COPD = chronic obstructive pulmonary disease; HR, hazard ratio; NSTEMI = non-ST-elevated myocardial infarction; STEMI = ST-elevated myocardial infarction.

4. DISCUSSION

In this nationwide population-based cohort and propensity score-matched study, AMI, STEMI, and NSTEMI patients with COPD had higher mortality rates than those without COPD. Guidelines suggest that post-AMI medications reduce the mortality rate in all patients with AMI, irrespective of COPD status. However, post-AMI medications are lesser administered to patients with a history of COPD. In contrast, medications for COPD may be misused in patients with AMI without COPD and underused in those with COPD. In patients with AMI, SABA, SAMA, and corticosteroids were associated with a higher mortality rate in all cohorts, whereas LABA, LAMA, and Xanthium were associated with a reduced mortality rate in AMI and NSTEMI cohorts.

AMI is associated with higher inhospital, short-term, and long-term mortality.11 The largest retrospective study conducted in the UK showed that both inhospital and 180-day mortality increased in patients with AMI and COPD.12 A systematic review and meta-analysis, including observational studies from Europe and the US, demonstrated that mortality risk is 26% higher in patients with AMI and COPD than in those with AMI without COPD.13 However, the heterogeneity of the meta-analysis was high (I2 = 74%), the number of patients included was <10 000, and the follow-up duration was variable (1-7 years). Our study used data from a nationwide, full-population database that provided long-term follow-up data. Our study suggests that COPD is an independent risk factor for long-term mortality in patients with AMI. This result was also observed in the STEMI and NSTEMI cohorts.

An analysis using the largest nationwide AMI registry in the US showed that STEMI was associated with higher mortality within 90 days than NSTEMI, but the difference was reduced after multivariant adjustment after 90 days.14 Our study demonstrated that patients with STEMI had lower long-term mortality than those with NSTEMI. This may be because patients with STEMI received more medications for secondary prevention and had a higher incidence of coronary artery interventions than those with NSTEMI (Table 1). STEMI has been associated with early mortality; however, our findings suggest that patients with NSTEMI require more aggressive evidence-based management than those with STEMI.

In the West, patients with AMI and COPD have received fewer evidence-based AMI medications.13–15 Our study revealed a similar trend of medication underusage, which is common in Taiwan. It is not surprising that the beta-blocker prescription rate is lower in patients with AMI and COPD. The use of beta-blockers in COPD raises concerns regarding acute exacerbation despite recent studies proving its safety.16 Underusing standard postmyocardial infarction medications has been shown to cause inferior outcomes.13,17

The present study showed that SABA and SAMA were associated with 30% and 20% higher mortality rates, respectively (Table 2). Intriguingly, no adverse effect on survival was observed using long-acting bronchodilators. A previous systematic review and meta-analysis investigating the association between inhaled bronchodilators and AMI revealed that tiotropium had a protective effect against myocardial infarction (26% risk reduction). Still, short-acting beta-2 agonists slightly increased the risk of myocardial infarction after the first inhalation.