American Journal of Respiratory and Critical Care Medicine

Rationale: Cardiovascular drugs may improve survival in chronic obstructive pulmonary disease (COPD). However, previous studies did not account for major sources of bias, and drug effects have not been evaluated in severe COPD.

Objectives: To estimate the time-dependent effects of cardiovascular drugs on survival in oxygen-dependent COPD, accounting for immortal and immeasurable time bias.

Methods: Prospective national study of patients starting long-term oxygen therapy for COPD in Sweden between 1 October 2005 and 30 June 2009. Effects on mortality were estimated using extended Cox regression adjusted for age, sex, PaO2, PaCO2, World Health Organization performance status, body mass index, comorbidity, and concomitant medications. Immortal and immeasurable time bias was addressed by analyzing all medications as time-dependent variables and accounting for hospitalized time, respectively.

Measurements and Main Results: Time-dependent effects of angiotensin-converting enzyme inhibitors or angiotensin receptor blockers, antiplatelet drugs, β-blockers, and statins on all-cause mortality were measured. Of the 2,249 included patients, 1,129 (50%) died under observation. No patient was lost to follow-up. The adjusted time-dependent model was compatible with reduced mortality for antiplatelet drugs (hazard ratio [HR], 0.86; 95% CI, 0.75–0.99; P = 0.030) and trends for angiotensin-converting enzyme inhibitors or angiotensin receptor blockers (HR, 0.90; 95% CI, 0.79–1.04; P = 0.166) and statins (HR, 0.86; 95% CI, 0.72–1.03; P = 0.105), whereas β-blockers increased mortality (HR, 1.19; 95% CI, 1.04–1.37; P = 0.010).

Conclusions: This study supports that antiplatelet drugs improve survival and β-blockers decrease survival in oxygen-dependent COPD.

Scientific Knowledge on the Subject

Cardiovascular drugs may improve survival in chronic obstructive pulmonary disease (COPD), but previous studies did not account for major sources of bias, and drug effects have not been evaluated in very severe COPD.

What This Study Adds to the Field

This prospective study supports a beneficial effect of antiplatelet drugs on survival and beneficial trends for angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, and statins. However, β-blockers may decrease survival in oxygen-dependent COPD.

As chronic obstructive pulmonary disease (COPD) is a leading cause of mortality worldwide, treatments that improve survival are needed (1).

Comorbidity is known to be highly prevalent in COPD (2) and is an important predictor of mortality (3, 4). A number of nonrandomized studies have suggested a beneficial effect on mortality for cardiovascular drugs: angiotensin-converting enzyme inhibitors (ACEI) (57), angiotensin receptor blockers (ARB) (5), antiplatelet drugs (7), β-blockers (712), and statins (57, 1315).

However, the surprisingly large effect sizes in previous studies have raised the concern that the findings might have been biased. According to a recent analysis (16), the apparent drug effects in several of the studies (5, 13) were explained by immortal time bias and immeasurable time bias (17, 18). These biases might in fact have affected most previous cardiovascular drug studies in COPD (57, 9, 11, 13).

Immortal time bias may arise when exposure is assessed after the start of follow-up. All patients who become exposed during follow-up must, by definition, have lived until the date of exposure, whereas nonexposed patients must not; this creates a spurious association between drug exposure and improved survival (16). Immeasurable time bias can arise when drug exposure cannot be measured during hospitalizations. Therefore, hospitalized patients with shorter survival are more likely to be misclassified as unexposed, which also creates a spurious association between drug exposure and improved survival (16). Methods to account for both types of bias have been described (16).

In addition, previous studies evaluated the effects of drug exposure defined at baseline only, which might be inappropriate as drug exposure is likely to vary over time. The effects of cardiovascular drugs have not been previously studied in patients with severe COPD, who have high mortality, including from cardiovascular disease (19).

We therefore conducted a national prospective study of the time-dependent effects of cardiovascular drugs on mortality in severe COPD, accounting for immortal and immeasurable time bias. Some of the results of this study were previously presented in the form of an abstract (20).

This was a national prospective multicenter study of patients aged 45 years or older registered in the national Swedevox Register, who started long-term oxygen therapy (LTOT) for physician-diagnosed COPD between 1 October 2005 and 30 June 2009. The Swedevox Register is administered by the Swedish Society of Respiratory Medicine and prospectively includes patients starting LTOT for COPD in Sweden since 1987, with a population-based coverage of some 85% (21).

Only the latest treatment episode was included for patients who had started LTOT more than once (n = 62). Exclusion criterion was a diagnosis of lung cancer before the date of starting LTOT (baseline). Data were collected at baseline on resting PaO2 and PaCO2 breathing air and during oxygen therapy, FEV1, FVC, body mass index (BMI), smoking history, and World Health Organization performance status (22).

Comorbidity and in-hospital time within 4 years before baseline were obtained from the National Patient Register for in- and outpatient care, which covers more than 99% of all hospitalizations since 1987 and about 80% of all hospital-based outpatient care in Sweden since 2001 (23). The 4-year assessment period assured that all patients had the same assessment time for comorbidity, as data on outpatient care were available after 2001. Comorbidities were coded according to the tenth revision of the International Classification of Disease (ICD) (24). Definitions of all diagnosis entities are found in Table E1 in the online supplement.

Data on medications were obtained from the Swedish Prescribed Drug Register, which includes all dispensed prescriptions in outpatient care in Sweden after 1 July 2005 (25). Medications were categorized according to the Anatomical Therapeutic Chemical Classification System codes (26), as shown in Table E2. Patients were followed until withdrawal of LTOT, death, or 31 December 2009, whichever came first. Vital status was obtained from the Swedish Causes of Death Register. The underlying cause of death was available between 2005 and 2008 (n = 758; 67% of total deaths).

All patients gave their informed consent to participate. The study was approved by all the relevant ethics committees in Sweden, the Swedish National Board of Health and Welfare, and the Data Inspection Board.

Statistical Analysis

Data were tabulated using frequencies and percentages for categorical variables, mean with SD, and median with range or interquartile range for continuous variables with normal and skewed distribution, respectively.

Drug exposure.

Dispensed prescribed drugs (prescriptions) were assessed during each full 91-day period (quarter) from the quarter before baseline until end of follow-up for each individual patient, as shown in Figure 1. A 91-day period was chosen, because the drugs in this study are usually prescribed for 3 months at a time. The probability of drug exposure for each quarter was set to 1 if there was at least one prescription during the quarter and set to 0 if there was no prescription during either this or the next quarter. To account for medications given in the hospital, the probability of exposure was set to the proportion of the current quarter spent in the hospital for quarters with no prescription but with a prescription during the next quarter. Exposure at baseline was defined as an exposure probability greater than 0 for the quarter before baseline. A more detailed description of the exposure assessment is provided in Section E1 in the online supplement.

Time-dependent model.

The exposures of primary interest were treatment with ACEIs or ARBs (merged into the category ACEI/ARB owing to few cases of exposed to ARB and presumed similar effects), antiplatelet drugs, β-blockers, and statins. Secondary hypotheses were those of treatment effects of the combination of a long-acting β2 agonist and an inhaled corticosteroid (LABA + ICS), oral glucocorticoids, and tiotropium.

Time-dependent drug effects on all-cause mortality were estimated using an extension of the Cox model (27). The time-varying covariate was set to the exposure probability for the 91-day period before the quarter of the event for each drug (Figure 1). This lag-time effect allowed for the time it takes for the cardiovascular medications to achieve their therapeutic effects, and the model assumed that the therapeutic effects remained for no more than 3 months after the last prescription.

Missing elements were imputed for PaO2 air (n = 289), PaCO2 air (n = 301), and BMI (n = 701), as detailed in Section E2 in the online data supplement. The model estimates were robust to the imputations. FEV1, smoking history, the number of previous hospitalizations, and the presence of diabetes mellitus were not included in the final analysis, as they had no significant independent effects and did not affect any of the multiple regression estimates.

In the final model, the drug effects were adjusted for baseline differences in age, sex, PaO2 air, PaCO2 air, BMI, World Health Organization performance status (22), and comorbidities: anemia, renal failure, and cardiovascular diseases (cerebrovascular disease, heart failure, hypertension, ischemic heart disease, peripheral artery disease, pulmonary embolism, or other circulatory disease).

Time-dependent drug effects, expressed as hazard ratios (HRs) with 95% confidence intervals (CIs), are interpreted as, at any given time, the hazard for a patient who was exposed during the previous quarter as compared with the hazard for a patient who was not exposed, adjusted for all other covariates.

Data management of dispensed prescriptions was performed using Mimer SQL version 10.1 (Mimer Information Technology AB, Uppsala, Sweden), and analyses were conducted with Stata version 11.1 (StataCorp LP; College Station, TX) and SAS version 9.2 (SAS Institute, Inc., Cary, NC).

In total, 2,249 patients, 1,328 (59%) women and 921 men, were included after excluding 2 (0.09%) patients due to data irregularities and 39 (2%) patients with a lung cancer diagnosis at baseline. No patient was lost to follow-up. The cohort generated 3,118.1 person-years at risk and was followed for a median 1.1 years (first quartile–third quartile, 0.6–2.0 yr). During this time, 1,129 (50%) patients died. The main causes of death were respiratory disease (68%), cardiovascular disease (20%), and cancer (6%).

Drug Exposure and Compliance

At baseline, ACEI/ARBs were used by 763 (34%) patients, antiplatelet drugs by 887 (39%), β-blockers by 829 (37%), and statins by 431 (19%). The number of patients treated with combinations of drugs and patient characteristics according to drug exposure at baseline are shown in Table 1. Patients treated with ACEI/ARB, antiplatelet drugs, β-blockers, or statins had higher BMI and more cardiovascular disease, diabetes mellitus, and renal failure than patients on none of these drugs (Table 1).

TABLE 1. PATIENT CHARACTERISTICS ACCORDING TO STUDY DRUG EXPOSURE AT BASELINE

CharacteristicAll PatientsACEI/ARBβ-BlockersStatinsNontreated
N (%)2,249 (100)763 (34)829 (37)431 (19)965 (43)
Age, yr74.7 ± 8.275.3 ± 7.775.6 ± 7.873.5 ± 7.174.1 ± 8.6
Women, n (%)1,328 (59)423 (55)461 (56)216 (50)611 (63)
PaO2 air, kPa6.5 ± 0.96.5 ± 0.96.5 ± 0.86.6 ± 0.86.6 ± 0.9
PaCO2 air, kPa6.3 ± 1.36.3 ± 1.36.3 ± 1.36.1 ± 1.36.2 ± 1.2
FEV1 % predicted37.7 ± 19.139.8 ± 18.940.5 ± 19.940.2 ± 20.735.9 ± 18.0
Known ever smoking, n (%)2,106 (93)726 (95)781 (94)408 (95)893 (93)
Body mass index, n (%)
 <18.5279 (12)60 (8)81 (10)42 (10)152 (16)
 18.5–24.9693 (31)203 (27)230 (28)115 (27)339 (35)
 25–29.9338 (15)125 (16)132 (16)78 (18)120 (12)
 ≥30235 (10)112 (15)113 (14)58 (13)64 (7)
WHO performance status, n (%)
 0132 (6)40 (5)37 (4)30 (7)63 (7)
 1881 (39)292 (38)321 (39)169 (39)392 (41)
 2714 (32)240 (31)268 (32)140 (32)302 (31)
 3292 (13)101 (13)116 (14)48 (11)113 (12)
 431 (1)10 (1)11 (1)3 (1)16 (2)
Hospitalizations*5 (2–8)5 (3–9)6 (3–9)5 (3–9)4 (2–8)
Anemia, n (%)196 (9)79 (10)95 (11)36 (8)70 (7)
Cardiovascular diagnoses, n (%)
 0755 (34)134 (18)116 (14)56 (13)510 (53)
 1823 (37)266 (35)330 (40)160 (37)318 (33)
 2449 (20)229 (30)245 (30)135 (31)98 (10)
 >2222 (10)134 (18)138 (17)80 (19)39 (4)
Diabetes mellitus, n (%)291 (13)159 (21)158 (19)120 (28)53 (5)
Renal failure, n (%)97 (4)55 (7)62 (7)34 (8)13 (1)
ACEI/ARB, n (%)763 (34)763 (100)415 (50)229 (53)0
Antiplatelet drugs, n (%)887 (39)402 (53)453 (55)321 (74)210 (22)
β-Blockers, n (%)829 (37)415 (54)829 (100)256 (59)0
LABA + ICS, n (%)1,562 (69)536 (70)570 (69)315 (73)657 (68)
Oral glucocorticoids, n (%)1,375 (61)437 (57)498 (60)282 (65)598 (62)
Statins, n (%)431 (19)229 (30)256 (31)431 (100)0
Tiotropium, n (%)1,165 (52)412 (54)427 (52)233 (54)482 (50)

Definition of abbreviations: ACEI = angiotensin-converting enzyme inhibitors; ARB = angiotensin receptor blockers; ICS = inhaled corticosteroids; LABA = long-acting β2 agonists; WHO = World Health Organization.

Data presented as mean ± SD unless otherwise specified. Percentages may not add up to 100 owing to rounding. Elements were missing for PaO2 air (n = 289), PaCO2 air (n = 301), FEV1 (n = 849), FVC (n = 863), body mass index (n = 701), and WHO performance status (n = 199). Hospitalizations and diagnoses were assessed within 4 years before baseline.

* Median (first quartile–third quartile).

The most commonly used ACEI/ARBs were enalapril and losartan, 98% of all β-blockers were cardioselective, and the antiplatelet and statin categories were dominated by aspirin and simvastatin, respectively. Dispensed drugs within each drug category are listed in Table E3. Of patients taking LABA or ICS, the majority (85%) used both drugs concomitantly; only few had monotherapy with LABA (5%) or ICS (10%). Table 2 shows changes in exposure status and drug adherence; of patients categorized as exposed to cardiovascular drugs at baseline, 10 to 14% changed status to unexposed during follow-up, and 5 to 10% of those unexposed at baseline were later exposed. During follow-up, patients were hospitalized for a median 7% (minimum, 0; maximum, 89%) of the time.

TABLE 2. CHANGES IN EXPOSURE STATUS FROM BASELINE DURING FOLLOW-UP IN 2,249 PATIENTS WITH OXYGEN-DEPENDENT CHRONIC OBSTRUCTIVE PULMONARY DISEASE

DrugFrom Exposed to Unexposed, n (% of Exposed)From Unexposed to Exposed, n (% of Unexposed)Continuously Exposed, n (% of Total)Median Adherence for Continuously Exposed (Min–Max), %
ACEI/ARB100 (14)102 (7)637 (28)89 (25–100)
Antiplatelet drugs93 (10)140 (10)759 (34)88 (18–100)
β-Blockers83 (10)128 (9)714 (32)92 (41–100)
LABA + ICS182 (12)173 (24)1,345 (60)87 (13–100)
Oral glucocorticoids354 (27)288 (31)980 (44)84 (15–100)
Statins55 (13)99 (5)365 (16)88 (38–100)
Tiotropium211 (18)160 (14)934 (42)89 (18–100)

Treatment status was measured per quarter (91-d period) during follow-up and adjusted for immeasurable (hospitalized) time. Baseline status denotes the exposure status of the quarter before inclusion. During follow-up, unexposed changed status at the first exposure, and exposed changed status when they were unexposed for all (and at least two) subsequent quarters. Compliance was calculated as the sum of the exposure probabilities for each full quarter observed, divided by the number of full quarters observed.

Effects on Mortality

The adjusted estimates were compatible with decreased mortality for antiplatelet drugs (HR, 0.86; 95% CI, 0.75–0.99; P = 0.030) and beneficial trends for ACEI/ARB (HR, 0.90; 95% CI, 0.79–1.04; P = 0.166) and statins (HR, 0.86; 95% CI, 0.72–1.03; P = 0.105), as shown in Table 3. In contrast, β-blockers were associated with increased mortality (HR, 1.19; 95% CI, 1.04–1.37; P = 0.010). In addition, oral glucocorticoids were significantly associated with higher mortality, and tiotropium tended to decrease mortality. The drug effects are depicted in Figure 2.

TABLE 3. TIME-DEPENDENT DRUG EFFECTS ON ADJUSTED ALL-CAUSE MORTALITY IN 2,249 PATIENTS WITH OXYGEN-DEPENDENT CHRONIC OBSTRUCTIVE PULMONARY DISEASE

ParameterHazard Ratio95% CIP Value
Age, per yr1.041.03–1.05<0.001
Woman0.770.67–0.87<0.001
BMI<0.001*
 <18.51.361.14–1.61<0.001
 18.5–24.9Ref
 25–29.90.720.60–0.87<0.001
 ≥300.790.64–0.990.044
WHO performance status<0.001*
 0Ref
 11.010.74–1.390.930
 21.471.07–2.000.017
 32.421.74–3.36<0.001
 43.281.98–5.42<0.001
 Missing1.370.96–1.960.082
PaO2 air, per 1 kPa0.910.85–0.980.009
PaCO2 air<0.001
Anemia1.241.02–1.500.034
Cardiovascular diagnoses<0.001*
 0Ref
 11.251.08–1.460.003
 21.411.18–1.69<0.001
 >21.391.11–1.750.005
Renal failure1.341.03–1.740.027
ACEI/ARB0.900.79–1.040.166
Antiplatelet drugs0.860.75–0.990.030
β-Blockers1.191.04–1.370.010
LABA + ICS0.980.86–1.120.790
Oral glucocorticoids1.531.35–1.73<0.001
Statins0.860.72–1.030.105
Tiotropium0.900.79–1.030.126

Definition of abbreviations: ACEI = angiotensin-converting enzyme inhibitors; ARB = angiotensin receptor blockers; CI = confidence interval; ICS = inhaled corticosteroids; LABA = long-acting β2 agonists; Ref = reference category; WHO = World Health Organization.

The hazard ratios for medications are interpreted as, at any given time, the hazard for a patient who was exposed during the 91-day period before the period of the event, as compared with the hazard for a patient who was unexposed, adjusted for all other covariates.

* Wald test of total significance for class variables with more than two categories.

Variable included as a second-degree polynomial, wherefore estimates are not reported.

There was no additional explanatory power in the model for interactions between ACEI/ARB and statins (P = 0.298), cardiovascular comorbidities and ACEI/ARB (P = 0.642), cardiovascular comorbidities and β-blockers (P = 0.909) or statins (P = 0.325), or between β-blockers and LABA + ICS or sex (P > 0.15 for both).

To explore the β-blocker effect, we estimated the time-dependent effects of β-blockers for subgroups of patients based on (1) the presence of cardiovascular disease, defined as at least one cardiovascular diagnosis at baseline; (2) exposure to LABA + ICS at baseline; and (3) sex. The negative impact of β-blockers tended to be stronger in patients without LABA + ICS, patients with cardiovascular disease, and in men (Figure E1). The increased mortality for β-blockers (HRs ≥ 1.20) was consistent across all subgroups except in women with LABA + ICS. However, there was no signs of interaction between β-blockers, sex, and LABA + ICS in the fully adjusted model (P = 0.782). Adjusting for ischemic heart disease in addition to the cardiovascular score did not change the β-blocker effect.

This study supports that antiplatelet drugs, and possibly ACEI/ARBs and statins, improve survival, whereas β-blockers decrease survival in patients with severe oxygen-dependent COPD.

This is the first study, to the authors’ knowledge, to evaluate the effects of cardiovascular drugs on mortality in very severe COPD and to account for immortal and immeasurable time bias as well as changes in drug exposure over time.

The improved survival for antiplatelet drugs and the beneficial trends for ACEI/ARBs and statins are consistent with studies of patients with less severe COPD (57, 11, 1315). van Gestel and colleagues reported that statins reduced mortality (HR, 0.67; 95% CI, 0.52–0.86) (10, 14). The study by van Gestel and colleagues is interesting because it was likely free from immortal and immeasurable time bias (14) but was a single-center study of a selected population of patients with COPD undergoing vascular surgery. Improved survival for statins was also found by Rutten and colleagues (HR, 0.83; 95% CI, 0.65–1.08), but no effect was found for ACEI/ARB (HR, 1.01; 95% CI, 0.84–1.21) (11). Short and colleagues recently reported a reduced mortality for ACEI/ARB (HR, 0.79; 95% CI, 0.72–0.88), antiplatelet drugs (HR, 0.80; 95% CI, 0.73–0.88), and statins (HR, 0.89; 95% CI, 0.81–0.97) (7). Compared with previous findings, the effect sizes in the present study were smaller, which might be because we accounted for immortal and immeasurable time bias.

Despite initial concerns, cardioselective β-blockers have in recent years been considered safe and effective to use in COPD based on studies of patients with mostly mild to moderate airflow limitation (712, 28, 29). A recent meta-analysis reported that β-blockers were associated with a lower relative risk of mortality of 0.69 (95% CI, 0.62–0.78) but also indicated the presence of publication bias; studies reporting nonbeneficial effects of β-blockers were less likely to be published (12). Our finding of increased mortality in oxygen-dependent COPD is in line with a previous report that the positive effect of β-blockers in milder forms of COPD is lost in patients with more severe respiratory impairment (29).

Strengths of the present study are that it included a large representative national sample of patients with severe oxygen-dependent COPD with complete follow-up. The study could be specifically designed to address immortal and immeasurable time bias, as complete nationwide data was available on hospitalizations and dispensed outpatient drugs. The validity of our findings is supported by the fact that we found effects of LABA + ICS and tiotropium that were compatible with those of the randomized TORCH and UPLIFT trials (30, 31), in contrast to the dramatic effect shown in previous nonrandomized studies (7, 11, 13).

There were two main limitations to our study design, the first being that dispensed prescriptions do not necessarily imply consumption. Adherence to the cardiovascular drugs was, however, generally high among users, and most exposed patients received regular prescriptions, implying that they actually took their drugs. The second limitation, common to all observational designs, is possible confounding by indication owing to the lack of randomization (32). Previous studies have tried to compensate for this by using propensity score methods, which model with or match on the conditional probability of receiving the treatment using a high-dimensional set of pretreatment characteristics (6, 7, 911, 33). However, the patient characteristics in this study, as in all previous studies in this field, were collected at study start and not at the start of each drug treatment, implying that the use of propensity score methods might be inappropriate. We addressed confounding by indication by restricting the analysis to patients with very severe COPD and adjusting for markers of disease severity, concomitant medication, and the presence of underlying indicatory diseases. These approaches have been found to be as effective as propensity score methods in terms of reducing confounding by indication, but the absence of residual bias needs to be confirmed by further observational studies and, ultimately, by randomized trials (34).

The mechanism between ACEI/ARB and statins, respectively, and reduced mortality could be their well-documented effects on cardiovascular mortality (35, 36). Our trend of reduced mortality independent of documented cardiovascular disease may be explained either by an effect on undiagnosed disease, because cardiac disease is difficult to identify in severe COPD (37), or by an effect on the excess risk of cardiovascular morbidity and mortality that seems to be a systemic consequence of COPD (38). ACEI/ARBs and statins may also have direct effects on the respiratory system. Animal models recently found effects of ARB on emphysema, exercise capacity, and lung function (39), and effects of simvastatin on smoke-induced pulmonary artery remodeling and emphysema (40). Treatment with an ARB was also shown to decrease the lung hyperinflation in patients with COPD in a placebo-controlled randomized trial (41).

A link between antiplatelet drugs and improved survival could be the systemic antithrombotic effect, because COPD is associated with increased platelet activation (42).

The association between oral glucocorticoids and increased mortality is in line with previous reports and might reflect a causal relation or an association between taking oral steroids and more severe underlying disease and exacerbations (4345).

Data are limited on adverse effect of β-blockers in severe COPD. A recent Cochrane review found that β-blockers had no negative effects on respiratory symptoms or FEV1 (46). However, this conclusion was based on a small number of relatively old studies with short follow-up, and these studies generally excluded patients with heart failure and other significant comorbidities (46, 47). In contrast, cardioselective β-blockers worsened the airway obstruction in patients with COPD in a randomized double-blind crossover trial (47, 48).

To further explore possible mechanisms behind the increased mortality, we studied the β-blocker effects in different patient subgroups. The adverse effect on survival was generally consistent across groups but tended to be lower for patients using LABA + ICS. This could indicate that long-acting bronchodilation counteracts some of the adverse respiratory effects of β-blockers (47), but subanalyses, including analyses of interaction, have low power and should be interpreted with caution. It cannot be precluded that β-blockers have adverse effects on lung function, airway responsiveness, or other parameters that could increase mortality in frail patients with very severe COPD.

In conclusion, the possible detrimental effect of β-blockers in patients with severe COPD is a novel and important finding that needs to be further validated. In addition, this study supports that antiplatelet drugs, and possibly ACEI/ARB and statins, have beneficial effects on survival in oxygen-dependent COPD.

The authors thank all physicians and nurses who included and cared for the patients.

1. Global Initiative for Chronic Obstructive Lung Disease (GOLD). Global strategy for the diagnosis, management and prevention of COPD. 2011 [accessed 2012 Aug 29]. Available from http://www.goldcopd.org/
2. Soriano JB, Visick GT, Muellerova H, Payvandi N, Hansell AL. Patterns of comorbidities in newly diagnosed COPD and asthma in primary care. Chest 2005;128:20992107.
3. Antonelli Incalzi R, Fuso L, De Rosa M, Forastiere F, Rapiti E, Nardecchia B, Pistelli R. Co-morbidity contributes to predict mortality of patients with chronic obstructive pulmonary disease. Eur Respir J 1997;10:27942800.
4. Mannino DM, Thorn D, Swensen A, Holguin F. Prevalence and outcomes of diabetes, hypertension and cardiovascular disease in COPD. Eur Respir J 2008;32:962969.
5. Mancini GB, Etminan M, Zhang B, Levesque LE, FitzGerald JM, Brophy JM. Reduction of morbidity and mortality by statins, angiotensin-converting enzyme inhibitors, and angiotensin receptor blockers in patients with chronic obstructive pulmonary disease. J Am Coll Cardiol 2006;47:25542560.
6. Mortensen EM, Copeland LA, Pugh MJV, Restrepo MI, de Molina RM, Nakashima B, Anzueto A. Impact of statins and ace inhibitors on mortality after COPD exacerbations. Respir Res 2009;10:45.
7. Short PM, Lipworth SIW, Elder DHJ, Schembri S, Lipworth BJ. Effect of beta blockers in treatment of chronic obstructive pulmonary disease: a retrospective cohort study. BMJ 2011;342:d2549.
8. Au DH, Bryson CL, Fan VS, Udris EM, Curtis JR, McDonell MB, Fihn SD. Beta-blockers as single-agent therapy for hypertension and the risk of mortality among patients with chronic obstructive pulmonary disease. Am J Med 2004;117:925931.
9. Dransfield MT, Rowe SM, Johnson JE, Bailey WC, Gerald LB. Use of beta blockers and the risk of death in hospitalised patients with acute exacerbations of COPD. Thorax 2008;63:301305.
10. van Gestel YRBM, Hoeks SE, Sin DD, Welten GMJM, Schouten O, Witteveen HJ, Simsek C, Stam H, Mertens FW, Bax JJ, et al.. Impact of cardioselective beta-blockers on mortality in patients with chronic obstructive pulmonary disease and atherosclerosis. Am J Respir Crit Care Med 2008;178:695700.
11. Rutten FH, Zuithoff NPA, Hak E, Grobbee DE, Hoes AW. Beta-blockers may reduce mortality and risk of exacerbations in patients with chronic obstructive pulmonary disease. Arch Intern Med 2010;170:880887.
12. Etminan M, Jafari S, Carleton B, FitzGerald JM. Beta-blocker use and COPD mortality: a systematic review and meta-analysis. BMC Pulm Med 2012;12:48.
13. Soyseth V, Brekke PH, Smith P, Omland T. Statin use is associated with reduced mortality in COPD. Eur Respir J 2007;29:279283.
14. van Gestel YRBM, Hoeks SE, Sin DD, Simsek C, Welten GMJM, Schouten O, Stam H, Mertens FW, van Domburg RT, Poldermans D. Effect of statin therapy on mortality in patients with peripheral arterial disease and comparison of those with versus without associated chronic obstructive pulmonary disease. Am J Cardiol 2008;102:192196.
15. Sheng X, Murphy MJ, MacDonald TM, Schembri S, Simpson W, Winter J, Winter JH, Wei L. Effect of statins on total cholesterol concentrations, cardiovascular morbidity, and all-cause mortality in chronic obstructive pulmonary disease: a population-based cohort study. Clin Ther 2012;34:374384.
16. Suissa S. Co-morbidity in COPD: the effects of cardiovascular drug therapies. Respiration 2010;80:37.
17. Suissa S. Immortal time bias in pharmaco-epidemiology. Am J Epidemiol 2008;167:492499.
18. Suissa S. Immeasurable time bias in observational studies of drug effects on mortality. Am J Epidemiol 2008;168:329335.
19. Ekstrom MP, Wagner P, Strom KE. Trends in cause-specific mortality in oxygen-dependent chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2011;183:10321036.
20. Ekstrom MP, Hermansson AB, Strom KE. Effect of cardiovascular drugs on mortality in severe chronic obstructive pulmonary disease: a time-dependent analysis [abstract]. Am J Respir Crit Care Med 2012;185:A2263.
21. Strom K, Boe J. A national register for long-term oxygen therapy in chronic hypoxia: preliminary results. Eur Respir J 1988;1:952958.
22. Oken MM, Creech RH, Tormey DC, Horton J, Davis TE, McFadden ET, Carbone PP. Toxicity and response criteria of the Eastern Cooperative Oncology Group. Am J Clin Oncol 1982;5:649655.
23. Swedish National Board of Health and Welfare. Quality and contents of the Swedish hospital discharge register. Stockholm: Swedish National Board of Health and Welfare. The Center for Epidemiology; 2008.
24. World Health Organization. Manual of the international statistical classification of diseases and health related problems. 10th revision. Geneva, Switzerland: World Health Organization; 1992.
25. Wettermark B, Hammar N, Fored CM, Leimanis A, Otterblad Olausson P, Bergman U, Persson I, Sundstrom A, Westerholm B, Rosen M. The new Swedish prescribed drug register–opportunities for pharmacoepidemiological research and experience from the first six months. Pharmacoepidemiol Drug Saf 2007;16:726735.
26. WHO Collaborating Centre for Drug Statistics Methodology. Guidelines for ATC classification and DDD assignment 2011. Oslo: Norwegian Institute of Public Health; 2010.
27. Kleinbaum DG, Klein M. Extension of the Cox proportional hazards model for time-dependent variables. In: , Kleinbaum DG, Klein M. Survival analysis, 2nd ed. New York: Springer-Verlag; 2005. pp. 211256.
28. Gottlieb SS, McCarter RJ, Vogel RA. Effect of beta-blockade on mortality among high-risk and low-risk patients after myocardial infarction. N Engl J Med 1998;339:489497.
29. Chen J, Radford MJ, Wang Y, Marciniak TA, Krumholz HM. Effectiveness of beta-blocker therapy after acute myocardial infarction in elderly patients with chronic obstructive pulmonary disease or asthma. J Am Coll Cardiol 2001;37:19501956.
30. Celli B, Decramer M, Kesten S, Liu D, Mehra S, Tashkin DP. Mortality in the 4-year trial of tiotropium (UPLIFT) in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2009;180:948955.
31. Calverley PM, Anderson JA, Celli B, Ferguson GT, Jenkins C, Jones PW, Yates JC, Vestbo J. Salmeterol and fluticasone propionate and survival in chronic obstructive pulmonary disease. N Engl J Med 2007;356:775789.
32. Strom BL, editor. Pharmacoepidemiology. West Sussex: John Wiley &Sons; 2005.
33. D’Agostino RB. Propensity score methods for bias reduction in the comparison of a treatment to a non-randomized control group. Stat Med 1998;17:22652281.
34. Bosco JLF, Silliman RA, Thwin SS, Geiger AM, Buist DSM, Prout MN, Yood MU, Haque R, Wei F, Lash TL. A most stubborn bias: no adjustment method fully resolves confounding by indication in observational studies. J Clin Epidemiol 2010;63:6474.
35. Yusuf S, Sleight P, Pogue J, Bosch J, Davies R, Dagenais G; The Heart Outcomes Prevention Evaluation Study Investigators. Effects of an angiotensin-converting-enzyme inhibitor, ramipril, on cardiovascular events in high-risk patients. N Engl J Med 2000;342:145153.
36. LaRosa J, He J, Vupputuri S. Effect of statins on risk of coronary disease: meta-analysis of randomized controlled trials. JAMA 1999;282:23402346.
37. Rutten FH, Cramer M-JM, Lammers J-WJ, Grobbee DE, Hoes AW. Heart failure and chronic obstructive pulmonary disease: an ignored combination? Eur J Heart Fail 2006;8:706711.
38. Sin DD, Wu L, Man SF. The relationship between reduced lung function and cardiovascular mortality: a population-based study and a systematic review of the literature. Chest 2005;127:19521959.
39. Raupach T, Luthje L, Kogler H, Duve C, Schweda F, Hasenfuss G, Andreas S. Local and systemic effects of angiotensin receptor blockade in an emphysema mouse model. Pulm Pharmacol Ther 2011;24:215220.
40. Wright JL, Zhou S, Preobrazhenska O, Marshall C, Sin DD, Laher I, Golbidi S, Churg AM. Statin reverses smoke-induced pulmonary hypertension and prevents emphysema but not airway remodeling. Am J Respir Crit Care Med 2011;183:5058.
41. Andreas S, Herrmann-Lingen C, Raupach T, Luthje L, Fabricius JA, Hruska N, Korber W, Buchner B, Criee CP, Hasenfuss G, et al.. Angiotensin II blockers in obstructive pulmonary disease: a randomised controlled trial. Eur Respir J 2006;27:972979.
42. Maclay JD, McAllister DA, Johnston S, Raftis J, McGuinnes C, Deans A, Newby DE, Mills NL, MacNee W. Increased platelet activation in patients with stable and acute exacerbation of COPD. Thorax 2011;66:769774.
43. Strom K. Survival of patients with chronic obstructive pulmonary disease receiving long-term domiciliary oxygen therapy. Am Rev Respir Dis 1993;147:585591.
44. Ringbaek TJ, Viskum K, Lange P. BMI and oral glucocorticoids as predictors of prognosis in COPD patients on long-term oxygen therapy. Chron Respir dis 2004;1:7178.
45. Schols AM, Wesseling G, Kester AD, de Vries G, Mostert R, Slangen J, Wouters EF. Dose dependent increased mortality risk in COPD patients treated with oral glucocorticoids. Eur Respir J 2001;17:337342.
46. Salpeter S, Ormiston T, Salpeter E. Cardioselective beta-blockers for chronic obstructive pulmonary disease. Cochrane Database Syst Rev 2005;CD003566.
47. Hawkins NM, Petrie MC, Macdonald MR, Jhund PS, Fabbri LM, Wikstrand J, McMurray JJV. Heart failure and chronic obstructive pulmonary disease the quandary of beta-blockers and beta-agonists. J Am Coll Cardiol 2011;57:21272138.
48. Dorow P, Thalhofer S, Bethge H, Disselhoff G, Wagner G. Long-term treatment of angina pectoris with bisoprolol or atenolol in patients with chronic obstructive bronchitis: a randomized, double-blind crossover study. J Cardiovasc Pharmacol 1990;16:S36S44.
Correspondence and requests for reprints should be addressed to Magnus Ekström, M.D., Ph.D., Department of Medicine, Blekinge Hospital, SE-37185, Karlskrona, Sweden. E-mail: .

Funded by the Research Council of Blekinge, the Swedish Heart-Lung Foundation, and the Swedish National Board of Health and Welfare.

Author Contributions: M.P.E. had full access to all the data in the study and takes full responsibility for the integrity of the data and the accuracy of the data analysis. Conception and design: M.P.E., A.B.H., K.E.S. Acquisition of data: M.P.E., K.E.S. Analysis and interpretation of data: M.P.E., A.B.H. Drafting the article: M.P.E., A.B.H. Revising it for important intellectual content and approval of the version to be published: M.P.E., A.B.H., K.E.S.

This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org

Originally Published in Press as DOI: 10.1164/rccm.201208-1565OC on January 17, 2013

Author disclosures

Related

No related items
American Journal of Respiratory and Critical Care Medicine
187
7

Click to see any corrections or updates and to confirm this is the authentic version of record