Chronic obstructive pulmonary disease (COPD) is associated with increased cardiovascular morbidity and mortality, yet the exact pathophysiological links remain unclear. Whether the presence and severity of COPD are associated with atrial or ventricular arrhythmias recorded on continuous electrocardiographic monitoring is unknown. We identified consecutive adult patients who underwent clinically indicated pulmonary function testing as well as 24-hour Holter monitoring at the Mayo Clinic, Rochester, from 2000 to 2009. Demographic data and relevant co-morbidities were gathered from the electronic medical record; severity of COPD was classified according to the GOLD classification, and arrhythmias were classified in concordance with the current clinical guidelines. From 7,441 patients who were included (age 64 ± 16 years, 49% woman, 92% Caucasian), COPD was diagnosed in 3,121 (41.9%). Compared with those without COPD, the presence and severity of COPD were associated with increased likelihood of atrial fibrillation/atrial flutter (AF/AFL; 23.3% vs 11.0%, respectively, p <0.0001), nonsustained ventricular tachycardia (NSVT; 13.0% vs 5.9%, respectively, p <0.0001), and sustained ventricular tachycardia (0.9% vs 1.6%, respectively, p <0.0001). COPD remained a significant predictor of AF/AFL and NSVT (p <0.0001 and p <0.0001, respectively) after adjusting for age, gender, tobacco use, obesity, hypertension, coronary artery disease, heart failure, diabetes, anemia, cancer, chronic kidney disease, and rate/rhythm control medications. In conclusion, the independent association between the presence and severity of COPD and arrhythmias (AF/AFL and NSVT) provides further insight into the markedly increased cardiovascular mortality of patients with COPD. Further studies should explore which anti-arrhythmic strategies would best apply to the patients with COPD.
In the United States, chronic obstructive pulmonary disease (COPD) ranks as the fourth leading cause of mortality, second leading cause of morbidity, and it remains the only common mortality etiology that continues to rise (COPD deaths increased by 58% from 1990 to 2010). Large prospective and retrospective studies of patients with COPD showed that the most common causes of their demise were cardiovascular events, and limited evidence suggests that some could be in part due to arrhythmias. Previous studies using continuous electrocardiographic (ECG) monitoring in patients with COPD were limited in scope (largest n = 75) and did not stratify COPD by its objectively documented severity. Knowing whether COPD is associated with arrhythmias could be crucial in appropriately directing our efforts aimed at reversing the alarming rise in COPD-related cardiovascular mortality in the United States and also around the World. We hypothesized that the most clinically significant arrhythmias (atrial fibrillation [AF], nonsustained ventricular tachycardia [NSVT], and sustained ventricular tachycardia [SustVT]) would be more commonly seen on 24-hour ECG Holter monitors of patients with COPD compared with patients without COPD, and that the severity of COPD would correlate with the prevalence of these arrhythmias.
Methods
This study protocol was approved by the Mayo Clinic Institutional Review Board. We retrospectively identified and searched the medical records at our institution for all unique adult patients who underwent clinically indicated 24-hour ECG Holter monitoring between the years 2000 and 2009. From this cohort, we identified those patients who performed pulmonary function testing (PFT) in our laboratory (spirometry variables necessary for inclusion: FEV1 and FVC). Relevant demographic variables, co-morbid conditions, and medication use at the time of Holter monitoring were identified from the electronic medical record, Holter monitor patient log book, and the international classification of diseases codes. Exclusion criteria were (1) incomplete medical record, (2) restrictive or nonspecific pulmonary disease identified on PFT, and (3) patient’s preference not to participate in research studies.
PFT was performed in the Mayo Clinic PFT Laboratory by a certified technician and the report then was read and confirmed by a board-certified pulmonary physician according to the current guidelines. The CareFusion MS Body-Diffusion system using whole-body plethysmography was utilized for the lung volume estimation. Severity of COPD was classified in accordance with the current guidelines in the following manner: (1) mild-to-moderate COPD defined as FEV1/FVC <0.7 and FEV1 percent predicted ≥0.5; (2) severe COPD defined as FEV1/FVC <0.7 and FEV1 percent predicted between 0.3 and 0.5 (0.3 ≤ FEV1 p < 0.5); and (3) very severe COPD defined as FEV1/FVC <0.7 and FEV1 percent predicted <0.3. Absence of COPD was defined as FEV1/FVC ≥0.7 and FEV1 percent predicted ≥0.8. Patients with restrictive or nonspecific pulmonary function test results were excluded.
Portable 3-channel Holter monitoring systems (Spacelabs Healthcare, Issaquah, Washington) were used for the 24 hours ECG recording, and the analysis was performed at the Mayo Clinic ECG laboratory by a certified ECG technician supervised by a cardiovascular physician (both blinded to the hypothesis of this study) according to the current guidelines. If the patients underwent 48-hour ECG Holter monitoring, we used only the first 24 hours of the recording. Ventricular tachycardia (VT) was defined as a sequence of ≥3 beats at a rate of >100 beats/min, which were ventricular in origin according to the clinical interpretation by the responsible board-certified cardiologist. VT that lasted <30 consecutive beats was termed NSVT, and VT lasting ≥30 consecutive beats was termed SustVT. Co-morbid conditions were determined from the electronic medical record and the International Classification of Diseases – 9th Revision coding. Medication use was retrieved from the patient’s log book at the time of the Holter monitoring and then grouped into the following categories: β blockers, calcium channel antagonists, amiodarone, and other antiarrhythmic medications (propafenone, flecainide, dofetilide, sotalol, procainamide, digitalis, mexiletine, and quinidine).
Statistical method included calculation of means and SDs for continuous variables and frequencies for categorical variables. The student t or Wilcoxon test was used for normally distributed or skewed variables, respectively. We performed chi-square or Fisher exact test to compare the difference among proportions. To determine the relation between the presence of COPD (dependent variable) and the occurrence of AF, NSVT, and SustVT as recorded by Holter monitoring (independent variables), we performed univariate analyses as well as multivariate logistic regression models (including assessments of potential confounding, colinearity, and effect modification). Variables of clinical interest were gradually introduced into the models, and the effect on the 95% confidence interval of the β coefficient related to the independent association between COPD and AF, NSVT, and SustVT, respectively, was noted. The final multivariate models (for each type of arrhythmia) were adjusted for age, sex, body mass index, nicotine dependence, coronary artery disease, hypertension, heart failure, anemia, diabetes mellitus, chronic kidney disease, and cancer. The data were analyzed using JMP version 9.0. Value <0.05 was considered statistically significant.
Results
We included 7,441 patients in this study (age 64 ± 16 years, 49% woman, 92% Caucasian) of whom 3,121 (41.9%) were diagnosed with COPD based on the PFT results. Characteristics of the study cohort stratified according to the presence and severity of COPD are listed in Table 1 . The Holter monitoring results (listed in Table 2 ) revealed that patients with COPD experienced significantly greater number of atrial premature complex (APC) and ventricular premature complex (VPC); this difference remained significant when applying previously used cut offs of ≥30 premature beats per hour for APCs and VPCs, respectively. The percentage of patients who were found to have AF during the 24-hour Holter recording period was significantly greater among those with COPD compared with patients without COPD (p <0.001), and a positive correlation between the severity of COPD and the preponderance of AF was found ( Figure 1 ). In a multivariate model, COPD remained a significant risk factor for the presence of AF on Holter monitoring (p <0.001, Figure 2 and Table 3 ). NSVT and SustVT were recorded more commonly in patients with COPD compared with patients with non-COPD (p <0.001 for NSVT, p = 0.012 for SustVT; Table 2 ). The frequency of NSVT and SustVT increased with worsening COPD severity ( Figure 1 ). In a multivariate analysis, COPD was found to be an independent risk factor for the occurrence of NSVT (p <0.001; Figure 2 and Table 3 ) but not for the occurrence of SustVT.
| Variable | COPD | p | |||
|---|---|---|---|---|---|
| No (n = 4,320) | Mild/Mod. (n = 2,239) | Severe (n = 698) | Very Severe (n = 184) | ||
| Age (years) | 59.8 ± 16.7 | 70.8 ± 11.6 | 71.6 ± 10.6 | 69.9 ± 11.9 | |
| Women | 2,533 (59%) | 846 (38%) | 239 (34%) | 54 (29%) | <0.0001 |
| Body mass index (kg/m 2 ) | 29.0 ± 6.2 | 28.7 ± 5.7 | 28.9 ± 6.6 | 26.9 ± 6.5 | 0.0078 |
| Forced expiratory volume in 1 second (percent predicted) | 0.97 ± 0.11 | 0.72 ± 0.14 | 0.41 ± 0.06 | 0.24 ± 0.04 | <0.0001 |
| Forced vital capacity (percent predicted) | 0.97 ± 0.11 | 0.88 ± 0.15 | 0.62 ± 0.13 | 0.49 ± 0.13 | <0.0001 |
| Forced expiratory volume in 1 second/forced vital capacity | 0.79 ± 0.05 | 0.63 ± 0.06 | 0.53 ± 0.10 | 0.40 ± 0.11 | <0.0001 |
| Smoker | 406 (9%) | 311 (14%) | 130 (19%) | 36 (20%) | <0.0001 |
| Coronary artery disease | 1,877 (43%) | 1,424 (64%) | 452 (65%) | 117 (64%) | <0.0001 |
| Heart failure | 894 (21%) | 826 (37%) | 343 (49%) | 108 (59%) | <0.0001 |
| Diagnosed hypertension | 2,799 (65%) | 1,741 (78%) | 529 (76%) | 136 (74%) | <0.0001 |
| Anemia ∗ | 1,628 (38%) | 1,024 (46%) | 338 (48%) | 94 (51%) | <0.0001 |
| Diabetes mellitus | 891 (21%) | 578 (26%) | 228 (33%) | 55 (30%) | <0.0001 |
| Chronic kidney disease | 512 (12%) | 434 (19%) | 164 (23%) | 45 (24%) | <0.0001 |
| Diagnosed cancer | 821 (19%) | 688 (31%) | 201 (29%) | 53 (29%) | <0.0001 |
| β-blocker use | 887 (21%) | 623 (28%) | 187 (27%) | 33 (18%) | <0.0001 |
| Calcium channel antagonist use | 240 (6%) | 168 (8%) | 66 (9%) | 23 (13%) | <0.0001 |
| Amiodarone use | 91 (2%) | 56 (3%) | 15 (2%) | 7 (4%) | 0.2703 |
| Other antiarrhythmic medication use | 305 (7%) | 240 (11%) | 103 (15%) | 31 (17%) | <0.0001 |
∗ Anemia was defined as present if noted in the electronic medical record or International Classification of Diseases – 9th Revision coding system.
| Variable | COPD | p | |||
|---|---|---|---|---|---|
| No (n = 4,320) | Mild/Mod. (n = 2,239) | Severe (n = 698) | Very Severe (n = 184) | ||
| Average heart rate (bpm) | 74.6 ± 12.1 | 73.1 ± 12.6 | 75.8 ± 14.5 | 79.8 ± 15.2 | 0.0618 |
| Minimum heart rate (bpm) | 66.7 ± 11.3 | 65.8 ± 11.7 | 69.0 ± 13.8 | 72.6 ± 14.5 | 0.5150 |
| Maximum heart rate (bpm) | 89.6 ± 14.1 | 86.6 ± 14.5 | 88.1 ± 15.8 | 92.6 ± 16.5 | <0.0001 |
| Mean ventricular premature complexes | 59.0 ± 203.8 | 94.0 ± 262.2 | 112.7 ± 261.5 | 134.8 ± 276.2 | <0.0001 |
| Number of ventricular premature complexes ≥ 30/hr | 880 (20.4%) | 641 (28.6%) | 276 (39.5%) | 80 (43.5%) | <0.0001 |
| Mean atrial premature complex beats | 36.0 ± 145.9 | 67.7 ± 234.9 | 73.8 ± 215.7 | 85.5 ± 355.1 | <0.0001 |
| Number of atrial premature complex ≥30/hr | 592 (13.7%) | 537 (24.0%) | 185 (26.5%) | 48 (26.1%) | <0.0001 |
| Number of sinus pauses | 0.5 ± 7.5 | 1.3 ± 10.2 | 1.2 ± 9.2 | 0.4 ± 4.7 | 0.0007 |
| Atrial fibrillation/flutter | 340 (11.0%) | 357 (21.8%) | 134 (26.0%) | 40 (31.8%) | <0.0001 |
| Nonsustained ventricular tachycardia | 255 (5.9%) | 256 (11.4%) | 112 (16.1%) | 39 (21.2%) | <0.0001 |
| Sustained ventricular tachycardia | 39 (0.9%) | 30 (1.3%) | 13 (1.9%) | 6 (3.3%) | 0.012 |
| Variable | COPD | |||
|---|---|---|---|---|
| No COPD (n = 4,320) | Mild/Mod. (n = 2,239) | Severe (n = 698) | Very Severe (n = 184) | |
| Atrial fibrillation/flutter | ||||
| Univariate | 1 ∗ | 2.27 (1.93–2.67) | 2.85 (2.27–3.57) | 3.78 (2.53–5.55) |
| Multivariate | 1 ∗ | 1.39 (1.17–1.66) | 1.56 (1.22–1.99) | 2.10 (1.36–3.18) |
| Nonsustained ventricular Tachycardia | ||||
| Univariate | 1 ∗ | 2.06 (1.72–2.47) | 3.05 (2.39–3.86) | 4.29 (2.91–6.18) |
| Multivariate | 1 ∗ | 1.34 (1.10–1.63) | 1.74 (1.34–2.24) | 2.24 (1.49–3.30) |
| Sustained ventricular tachycardia | ||||
| Univariate | 1 ∗ | 1.49 (0.92–2.40) | 2.08 (1.06–3.82) | 3.70 (1.39–8.23) |
| Multivariate | 1 ∗ | 0.92 (0.55–1.52) | 1.11 (0.55–2.10) | 1.80 (0.66–4.14) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
