The authors have declared that no competing interests exist.
Conceived and designed the experiments: MW MB AH FR MLDB AdB RK HT. Performed the experiments: MB JB AB PS RK. Analyzed the data: MW MB. Contributed reagents/materials/analysis tools: PS. Wrote the paper: MW MB MLDB HT.
We aimed to determine whether (1) patients with obstructive pulmonary disease (OPD) have an increased risk of sudden cardiac arrest (SCA) due to ventricular tachycardia or fibrillation (VT/VF), and (2) the SCA risk is mediated by cardiovascular risk-profile and/or respiratory drug use.
A community-based case-control study was performed, with 1310 cases of SCA of the ARREST study and 5793 age, sex and SCA-date matched non-SCA controls from the PHARMO database. Only incident SCA cases, age older than 40 years, that resulted from unequivocal cardiac causes with electrocardiographic documentation of VT/VF were included. Conditional logistic regression analysis was used to assess the association between SCA and OPD. Pre-specified subgroup analyses were performed regarding age, sex, cardiovascular risk-profile, disease severity, and current use of respiratory drugs.
A higher risk of SCA was observed in patients with OPD (n = 190 cases [15%], 622 controls [11%]) than in those without OPD (OR adjusted for cardiovascular risk-profile 1.4 [1.2–1.6]). In OPD patients with a high cardiovascular risk-profile (OR 3.5 [2.7–4.4]) a higher risk of SCA was observed than in those with a low cardiovascular risk-profile (OR 1.3 [0.9–1.9]) The observed SCA risk was highest among OPD patients who received short-acting β2-adrenoreceptor agonists (SABA) or anticholinergics (AC) at the time of SCA (SABA OR: 3.9 [1.7–8.8], AC OR: 2.7 [1.5–4.8] compared to those without OPD).
OPD is associated with an increased observed risk of SCA. The most increased risk was observed in patients with a high cardiovascular risk-profile, and in those who received SABA and, possibly, those who received AC at the time of SCA.
Sudden cardiac arrest (SCA) most often causes sudden death and is the most common direct cause of death in Western
Reports on SCA often use a practical but inaccurate definition of sudden death: witnessed natural death <1 hour of onset of acute symptoms, or unwitnessed unexpected death of someone seen in a stable medical condition <24 hours previously.
The AmsteRdam REsuscitation Study (ARREST) was conducted according to the principles expressed in the Declaration of Helsinki. Written informed consent was obtained from all participants who survived SCA. The Ethics Committee of the Academic Medical Center Amsterdam approved the use of data from patients who did not survive SCA, and approved this study.
We performed a community-based case-control study. Cases were SCA patients from the ARREST database. Each case was matched to five controls without SCA by age, sex and index date (date of SCA in cases) drawn from the PHARMO record linkage system (
ARREST is specifically designed to study the causes and outcome of SCA in the community (out-of-hospital). All individuals who suffer SCA in the North Holland province of the Netherlands (>2.4 million inhabitants) are included. The ARREST study protocol is described in detail elsewhere.
The PHARMO database includes drug-dispensing records from community pharmacies of >3 million community-dwelling inhabitants in the Netherlands. The catchment area covers about 12% of the total population of the Netherlands, and is representative of the total population. Since nearly all patients in the Netherlands are registered at a single community pharmacy, independent of prescriber, pharmacy records are essentially complete.
Because it is virtually impossible to disentangle the effects of OPD from those of respiratory drugs, as nearly all OPD patients use such drugs, and those more severely affected generally use more drugs (often from multiple drug classes),
As the association between OPD and SCA may be confounded by patient and other characteristics associated with both the presence of OPD and the risk of SCA, we studied the influence of various covariates on the calculated associations. Potential confounders studied were high cardiovascular risk-profile, diabetes mellitus, and current use of antiarrhythmic drugs or non-antiarrhythmic QTc prolonging drugs. High cardiovascular risk-profile was defined as the use of any of the following drugs within 6 months before index date: β-adrenoreceptor blockers, calcium channel antagonists, angiotensin-converting enzyme inhibitors, diuretics, angiotensin-II receptor blockers, nitrates, platelet aggregation inhibitors, and/or statins. Diabetes mellitus was defined by use of anti-diabetics within six months before index date. Anti-arrhythmic drugs were Vaughan-Williams class I or III antiarrhythmic drugs (
Differences in baseline characteristics were examined with chi-square tests or t-tests. Conditional logistic regression analysis was used to examine the association between SCA and OPD, with adjustment for three sets of confounders: 1) all potential confounders, 2) all covariates that were univariately associated with SCA (p<0.05), 3) all covariates that were univariately associated with SCA and that changed the beta-coefficient of the association between OPD and SCA by ≥5%. As logistic regression analyses were performed, odds ratios were calculated, which can be interpreted as a risk ratio in case of adequate sampling of controls from the study base. Stratified analyses were performed regarding age category, sex, and cardiovascular risk-profile. The presence of interaction on a multiplicative scale between OPD and cardiovascular risk-profile, was estimated by including the cross-product of the two factors as a variable in the model. The presence of interaction on an additive scale between OPD and high cardiovascular risk-profile was estimated by determining the synergy index.
During the study period, 3821 instances of cardiac arrest were recorded, of which 1875 cases had ECG-documented VT/VF and were aged >40 years. We excluded 565 patients (
The source region had a population of 2 426 097 people in 2007 [Netherlands Statistics.
Baseline characteristics | Cases | Controls | p-value |
N = 1310 | N = 5793 | ||
Mean age (years, standard deviation) | 67.1 (12.2) | 67.0 (12.2) | n/a |
Age group in years | |||
<65 | 546 (42%) | 2421 (42%) | |
≥65 | 764 (58%) | 3372 (58%) | n/a |
Male sex | 1015 (78%) | 4502 (78%) | n/a |
Comorbidities | |||
High cardiovascular risk-profile |
929 (71%) | 3049 (53%) | <0.001 |
Diabetes mellitus |
232 (18%) | 621 (11%) | <0.001 |
Current use of concomitant medication |
|||
Antiarrhythmic drugs |
60 (4%) | 182 (3%) | 0.009 |
β-adrenoreceptor blockers | 463 (35%) | 1134 (20%) | <0.001 |
Non-antiarrhythmic QT prolonging drugs class 1 | 21 (2%) | 64 (1%) | 0.134 |
Non-antiarrhythmic QT prolonging drugs class 2 | 35 (3%) | 151 (3%) | 0.891 |
Obstructive pulmonary disease | 190 (15%) | 622 (11%) | <0.001 |
Current β-adrenoreceptor blocker-use in OPD patients | 70 (37%) | 120 (19%) | <0.001 |
Current use of inhaled respiratory drugs |
|||
Inhaled short-acting β2-adrenoreceptor agonists | 62 (5%) | 51 (0.9%) | <0.001 |
Inhaled long-acting β2-adrenoreceptor agonists | 78 (6%) | 127 (2%) | <0.001 |
Inhaled anticholinergics | 78 (6%) | 102 (2%) | <0.001 |
Inhaled corticosteroids | 97 (7%) | 205 (4%) | <0.001 |
Other drugs used to treat OPD | |||
Systemic β2-adrenoreceptor agonists |
3 (0.2%) | 4 (0.1%) | 0.096 |
Xanthines |
10 (0.8%) | 31 (0.5%) | 0.325 |
Chronically used systemic corticosteroids |
18 (1.4%) | 93 (1.6%) | 0.542 |
Data are number (%) unless otherwise indicated. OPD: obstructive pulmonary disease.
Drug use at index date.
Drug use at index date, or within six months prior to index date.
Use of systemic corticosteroids with a duration of 90 days or more.
Use of any of the following drugs: β-adrenoreceptor blockers, calcium channel antagonists, angiotensin converting enzyme inhibitors, diuretics, angiotensin-II receptor blockers, nitrates, platelet aggregation inhibitors, and statins, within six months prior to index date.
Use of anti-diabetics within six months prior to index date.
Class I and III antiarrhythmic drugs and non-antiarrhythmic drugs with (possible) risk of QT prolongation (
OPD was more prevalent in cases (15%) than controls (11%, p<0.001). OPD was independently associated with an increased observed risk of SCA (crude OR 1.4 [1.2–1.7]). The three different models used to adjust for confounding resulted in similar ORs (
Outcome | UnadjustedOR (95%CI) | AdjustedOR |
AdjustedOR |
AdjustedOR |
Obstructive pulmonary disease | 1.4 (1.2–1.7) | 1.4 (1.1–1.6) | 1.4 (1.1–1.6) | 1.4 (1.2–1.6) |
High cardiovascular risk-profile |
2.5 (2.2–2.9) | 2.3 (2.0–2.7) | 2.3 (2.0–2.7) | 2.5 (2.2–2.9) |
Diabetes mellitus |
1.8 (1.5–2.1) | 1.5 (1.2–1.7) | 1.5 (1.2–1.7) | |
Use of antiarrhythmic drugs |
1.5 (1.1–2.0) | 1.2 (0.9–1.6) | 1.2 (0.9–1.6) | |
Non-antiarrhythmic QT prolonging drugs class 1 |
1.4 (0.8–2.3) | 1.2 (0.7–2.0) | ||
Non-antiarrhythmic QT prolonging drugs class 2 |
1.0 (0.7–1.5) | 1.0 (0.7–1.4) |
CI: confidence interval, OR: odds ratio.
Adjusted for all potential confounders.
Adjusted for all covariates that were univariately associated with sudden cardiac arrest.
Adjusted for all covariates that were univariately associated with sudden cardiac arrest and changed the beta with at least 5%.
Use of any of the following drugs: β-adrenoreceptor blockers, calcium channel antagonists, angiotensin converting enzyme inhibitors, diuretics, angiotensin-II receptor blockers, nitrates, platelet aggregation inhibitors, and/or statins, within six months prior to index date.
Use of anti-diabetics within six months prior to index date.
Class I and III antiarrhythmic drugs and non-antiarrhythmic drugs with (possible) risk of QT prolongation. (
Stratification according to sex showed a stronger effect of OPD in women (ORadj 1.8 [1.3–2.6]) than in men (ORadj 1.3 [1.03–1.6],
Outcome | Cases | Controls | Unadjusted | Adjusted OR |
||
N = 1310 | N = 5793 | OR (95% CI) | ||||
<65 with OPD | 58/546 (11%) | 163/2421 (7%) | 1.6 (1.2–2.3) | 1.6 (1.2–2.3) | ||
≥65 with OPD | 132/764 (17%) | 459/3372 (14%) | 1.3 (1.1–1.6) | 1.3 (1.03–1.6) | ||
Women with OPD | 51/295 (17%) | 124/1291 (10%) | 2.0 (1.3–2.8) | 1.8 (1.3–2.6) | ||
Men with OPD | 139/1015 (14%) | 498/4502 (11%) | 1.3 (1.04–1.6) | 1.3 (1.03–1.6) | ||
No OPD | Low risk profile | 342 (26%) | 2505 (43%) | Reference | n/a | |
High risk profile | 778 (59%) | 2666 (46%) | 2.5 (2.1–2.9) | n/a | ||
OPD | Low risk profile | 39 (3%) | 239 (4%) | 1.3 (0.9–1.9) | n/a | |
High risk profile | 151 (12%) | 383 (7%) | 3.5 (2.7–4.4) |
n/a |
Data are number (%). CI: confidence interval, CVD: cardiovascular disease, N: number, n/a: not applicable, OPD: obstructive pulmonary disease, OR: odds ratio.
Use of β- adrenoreceptor blockers, calcium channel antagonists, angiotensin converting enzyme inhibitors, diuretics, angiotensin-II receptor blockers, nitrates, platelet aggregation inhibitors, and/or statins within six months prior to index date.
Adjusted for cardiovascular risk profile.
Interaction on a multiplicative scale: OR 1.1 (0.7–1.6), on an additive scale: synergy index 1.4 (0.7–2.6).
When we studied cardiovascular risk-profile in detail, we found that SCA risk was observed to be most elevated in OPD patients with a high cardiovascular risk-profile (OR 3.5 [2.7–4.4]), and less so in patients without OPD, but with a high cardiovascular risk-profile (OR 2.5 [2.1–2.9]) or in OPD patients with a low cardiovascular risk-profile (OR 1.3 [0.9–1.9]). No significant interaction between cardiovascular risk-profile and OPD was observed on a multiplicative scale, nor on an additive scale. The observed SCA risk increased in parallel with OPD severity: compared to patients without OPD, the observed SCA risk was more elevated in patients with very severe OPD (ORadj 1.8 [1.2–2.7]) than in patients with moderate OPD (ORadj 1.4 [1.1–1.7],
Outcome | Cases | Controls | Unadjusted OR (95% CI) | Adjusted OR |
|
N = 1310 | N = 5793 | ||||
No OPD | 1120 (86%) | 5171 (89%) | Reference | Reference | |
OPD | Mild (0 drugs) | 12 (1%) | 61 (1%) | 0.9 (0.5–1.7) | 0.9 (0.5–1.7) |
Moderate (1–2 drugs) | 99 (8%) | 334 (6%) | 1.4 (1.1–1.7) | 1.4 (1.1–1.7) | |
Severe (3 drugs) | 45 (3%) | 147 (3%) | 1.4 (1.01–2.0) | 1.3 (0.9–1.9) | |
Very severe (>3 drugs) | 34 (3%) | 80 (1%) | 1.9 (1.3–2.9) | 1.8 (1.2–2.7) | |
No OPD | 1120 (86%) | 5171 (89%) | Reference | Reference | |
OPD | No SABA, AC, LABA or ICS | 52 (4%) | 358 (6%) | 0.7 (0.5–0.9) | 0.7 (0.5–0.9) |
SABA only | 12 (0.9%) | 13 (0.2%) | 4.1 (1.9–9.0) | 3.9 (1.7–8.8) | |
LABA only | 2 (0.2%) | 5 (0.1%) | 1.7 (0.3–9.0) | 1.8 (0.3–9.2) | |
AC only | 19 (2%) | 30 (0.5%) | 2.8 (1.6–5.0) | 2.7 (1.5–4.8) | |
ICS only | 11 (0.8%) | 78 (1.3%) | 0.6 (0.3–1.2) | 0.7 (0.4–1.3) | |
SABA+AC | 8 (0.6%) | 10 (0.2%) | 3.5 (1.4–8.9) | 2.6 (1.02–6.7) | |
ICS+LABA | 26 (2%) | 57 (1.0%) | 2.0 (1.3–3.3) | 2.0 (1.2–3.2) | |
SABA+LABA and/or ICS | 14 (1%) | 11 (0.2%) | 5.5 (2.5–12.1) | 5.3 (2.4–12.0) | |
AC+LABA and/or ICS | 23 (2%) | 46 (0.8%) | 2.3 (1.4–3.8) | 2.0 (1.2–3.4) | |
SABA+AC+LABA and/or ICS | 23 (2%) | 14 (0.2%) | 7.9 (3.9–16.0) | 7.6 (3.7–15.6) |
Data are number (%) or odds ratios (95% confidence interval).
ATC: Anatomical Therapeutic Chemical classification system, AC: anticholinergics, CI: confidence interval, ICS: inhaled corticosteroids, LABA: long-acting β2-adrenoreceptor agonists, N: number, OPD: obstructive pulmonary disease, OR: odds ratio, SABA: short-acting β2-adrenoreceptor agonists.
Adjusted for concomitant cardiovascular disease.
Number of different respiratory drugs used (ATC code R03) in the six-month period before index date. Patients with mild OPD are patients who received at least two prescriptions of any drug ATC code R03 (drugs for obstructive airway diseases), within one year prior to the index date, but who did not use any of these drugs in the six-month period before index date.
Analysis of current use of respiratory drug revealed that increased SCA risk was associated with the use of SABA only (ORadj 3.9 [1.7–8.8]) or AC only (ORadj 2.7 [1.5–4.8]), but not ICS only, while use of LABA only was too rare to draw any conclusions (
This is the first study to show that OPD is associated with a 40% increased risk of ECG-confirmed SCA. Multiple analytical approaches were applied to account as much as possible for confounding. These analyses consistently demonstrated a statistically significantly increased observed risk of SCA in OPD patients, with an OR of 1.4. This provides some evidence of the robustness of our findings, but residual confounding cannot be completely ruled out, because of potential misclassification in some of our measurements. The increase in observed SCA risk is most pronounced in OPD patients with a high cardiovascular risk-profile. Use of SABA increases SCA risk among OPD patients, particularly in patients with a high cardiovascular risk-profile: in these patients a six-fold increased SCA risk was observed compared to patients without OPD.
The observed increased risk of SCA in OPD patients may be, in part, due to the higher prevalence of concomitant cardiovascular disease and cardiac arrhythmias, as these are risk factors for VT/VF. However, adjusting our analyses for cardiovascular risk-profile, current use of anti-arrhythmic drugs (as a proxy for pre-existing cardiac arrhythmias), and QT prolonging drugs (which may evoke cardiac arrhythmias) did not alter the OR for SCA risk.
Still, the strongest association with SCA was observed in OPD patients with a high cardiovascular risk-profile, particularly, in those who used SABA. Other shared risk factors for OPD and cardiovascular disease may also play a role, but we were unable to assess them. Smoking, a common cause of COPD, is clearly also associated with ischemic heart disease.
To further support the role of OPD in SCA risk, we found that more severe OPD was more strongly associated with SCA risk than mild OPD. Possibly, chronic hypoxemia, more likely to occur in those with more severe OPD, may contribute to the development of cardiac arrhythmias by increasing resting heart rates, as increased heart rate is associated with increased mortality in both the general population as well as in patients with COPD.
Interestingly, although not statistically significantly, we found a stronger association between OPD and SCA in women than in men. This contrasts with the general population, where the observed SCA risk among women is only a third of that in men.
We observed that, in the studied population, patients who received SABAs at the time of SCA, especially when combined with other respiratory drugs, had a higher SCA risk than patients who received other respiratory drugs. For LABAs, this association was less clear. The association between SABA use and SCA may be explained as follows. β2-adrenoreceptor agonists act on the β2-adrenergic receptors of bronchial smooth muscle, leading to dilatation of the bronchi, which results in relief of symptomatic wheeze and dyspnoea, and improvement of lung function.
Consequently, soon after their availability in the 1960s, concerns have emerged about potentially serious adverse effects of β2-adrenoreceptor agonists on the heart, although the currently available studies and meta-analyses yield conflicting results. The meta-analysis of Salpeter
In accordance with previous studies which showed that AC use was associated with an increased risk of cardiovascular death,
A major strength of our study is that ARREST was specifically designed to study the determinants of SCA. This ensured that SCA diagnosis was accurate. SCA was validated by the presence of VT/VF on the ECG. This is especially important in OPD patients, because sudden death caused by cardiac arrest may easily be confused with sudden death caused by respiratory failure.
A limitation is that, as this is an observational study, we were unable to completely distinguish between the effects of OPD
We observed that the overall risk of SCA is 40% higher in patients with OPD than in patients without OPD, and we were able to identify subgroups of OPD patients in whom this risk was most elevated: those with a high cardiovascular risk-profile, those who receive SABA, and possibly those who receive AC at the time of SCA. Our findings may provide the basis for refinements in treatment strategies for OPD patients. We therefore would recommend a prospective trial to evaluate the effectiveness of more integrated pulmonary and cardiovascular care (e.g., investigations to detect previously unrecognized cardiovascular disease) in these high-risk patients.
(DOCX)
The authors greatly appreciate the contributions of Paulien Homma, Michiel Hulleman, Esther Landman and Renate van der Meer to the data collection, data entry, and patient follow-up. We are greatly indebted to all participating EMS dispatch centers and ambulance services for their cooperation and support. We thank all the students at the University of Amsterdam who helped collect the data of the onsite AEDs and patient’s pharmacies.